Word Spotting A New Approach to Indexing Handwriting
Word Spotting:A New Approach to Indexing Handwriting R.Manmatha,Chengfeng Han and E.M.Riseman,
Center for Intelligent Information Retrieval,
Computer Science Department,
University of Massachusetts,Amherst,MA-01003.
manmatha@https://www.wendangku.net/doc/fe983486.html,
Abstract
There are many historical manuscripts written in a sin-gle hand which it would be useful to index.Examples in-clude the W.B.DuBois collection at the University of Mas-sachusetts and the early Presidential libraries at the Library of Congress.Since Optical Character Recognition(OCR) does not work well on handwriting,an alternative scheme based on matching the images of the words is proposed for indexing such texts.
The current paper deals with the matching aspects of this process.Two different techniques for matching words are discussed.TheTrst method matches words assuming that the transformation between the words may be mod-elled by a translation(shift).The second method matches words assuming that the transformation between the words may be modelled by an afTne transform.
Experiments are shown demonstrating the feasibility of the approach for indexing handwriting.
The method should also be applicable to retrieving pre-viously stored material from personal digital assistants (PDAs).
1Introduction
Text has always been the primary source of informa-tion in conventional,and more recently digital libraries.If the text is in machine readable form(ASCII),it can be indexed using standard text retrieval engines.However, much of the text in both historical and even current col-lections is contained in paper documents.One solution is to use Optical Character Recognition(OCR)to convert scanned paper documents into ASCII.Existing OCR tech-nology works well with standard machine printed fonts against clean backgrounds.It works poorly if the originals are of poor quality or if the text is handwritten.We propose an alternative solution for indexing handwritten text when a large corpus of texts written by a single person exists.
In the context of digital libraries,the problem being ad-dressed in this paper is primarily related to the indexing This research was supported by the NSF Center for Intelligent Infor-mation Retrieval and by ARPA grant number N66001-94-D-6054.of historical manuscripts.These manuscripts are largely written in a single hand.Examples include the collected works of W.E.B.Du Bois,the African American civil rights leader,and Margaret Sanger,a pioneer in birth con-trol,which are stored at archives at the University of Mas-sachusetts and Smith College respectively.
Such manuscripts are valuable resources for scholars as well as others who wish to consult the original manuscripts and considerable effort has gone into manually producing indices for them.For example,a substantial collection of Margaret Sanger’s work has been recently put on microTlm (see https://www.wendangku.net/doc/fe983486.html,/Sanger/SangBase.HTM)with an item by item index.The indexing scheme proposed here will help in the creation and production of indices and con-cordances for such archives.
Since the document is written by a single person,the assumption is that the variation in the word images will be small.The proposed solution willTrst segment the page into words and then match the actual word images against each other to create equivalence classes.Each equivalence class will consist of multiple instances of the same word. Each word will have a link to the page it came from.The number of words in each equivalence class will be tabu-lated.Those classes with the largest numbers of words will probably be stopwords,i.e.conjunctions such as“and”or articles such as“the”.Classes containing stopwords are eliminated(since they are not very useful for indexing).A list is made of the remaining classes.This list is ordered occuring to the number of words contained in them.The user provides ASCII equivalents for a representative word in each of the top m(say m=2000)classes.The words in these classes can now be indexed.This technique will be called“word spotting”as it is analogous to“word spotting”in speech processing[4].
The proposed solution completely avoids machine recognition of handwritten words as this is a difTcult task [8].Robustness is achieved compared to OCR systems for two reasons:
1.Matching is based on entire words.This is in con-
trast to conventional OCR systems which essentially recognize characters rather than words.
2.Recognition is avoided.Instead a human is placed in
the loop when ASCII equivalents of the words must be provided.
The present paper deals with the matching aspects of the problem(for a discussion of page segmentation into words, see[7]).A future paper will deal with the rest of the sys-tem.The matching phase of the problem is expected to be the most difTcult part of the problem.This is because un-like machine fonts,there is some variation in even a single person’s handwriting.This variation is difTcult to model. Figure(1)shows two examples of the word“Lloyd”writ-ten by the same person.The last image is produced by XOR’ing these two images.The white areas in the XOR image indicate where the two versions of“Lloyd”differ. This result is not unusual.In fact,the differences are some-times even larger.
Figure1:Two examples of the word“Lloyd”and the XOR image
In this paper,two different matching techniques are dis-cussed.TheTrst,based on Euclidean distance mapping [2],assumes that the deformation between words can be modelled by a translation(shift).The second,based on an algorithm by Scott and Longuet Higgins[9]models the transformation between words using an afTne transform. 2Prior Work
The traditional approach to indexing documents in-volvesTrst converting them to ASCII and then using a text based retrieval engine[10].Scanned documents printed in standard machine fonts against clean backgrounds can be converted into ASCII using an OCR[1].However,hand-writing is much more difTcult for OCRs to handle because of the wide variability present in handwriting(not only is there variability between writers,but a given person’s writ-ing also varies).
Image matching of words has been used to recognize words in documents which use machine fonts[5].Recog-nition rates are much higher than when the OCR is used directly[5].Machine fonts are simpler to match than hand-written fonts since the variation is much smaller;multiple instances of a given word printed in the same font are iden-tical except for noise.In handwriting,however,multiple instances of the same word on the same page by the same writer show variations.TheTrst two pictures in Figure1 are two identical words from the same document,written by the same writer.It may thus be necessary to account for these variations.
3Outline of Algorithm
1.A scanned greylevel image of the document is ob-
tained.
2.The image isTrst reduced by half by gaussianTltering
and subsampling.
3.The reduced image is then binarized by thresholding
the image.
4.The binary image is now segmented into words.this
is done by a process of smoothing and thresholding (see[7]).
5.A given word image(i.e.the image of a word)is used
as a template.and matched against all the other word images.This is repeated for every word in the doc-ument.The matching is done in two phases.First, the number of words to be matched is pruned using the areas and aspect ratios of the word images-the word to be matched cannot have an area or aspect ratio which is too different from the template.Next,the ac-tual matching is done by using a matching algorithm.
Two different matching algorithms are tried here.One of them only accounts for translation shifts,while the other accounts for afTne matches.The matching di-vides the word images into equivalence classes-each class presumably containing other instances of the same word.
6.Indexing is done as follows.For each equivalence
class,the number of elements in it is counted.The top n equivalence classes are then determined from this list.The equivalence classes with the highest number of words(elements)are likely to be stopwords(i.e.
conjunctions like‘and’,articles like‘the’,and prepo-sitions like‘of’)and are therefore eliminated from further consideration.Let us assume that of the top n,m are left after the stopwords have been eliminated.
The user then displays one member of each of these m equivalence classes and assigns their ASCII interpre-tation.These m words can now be indexed anywhere they appear in the document.
We will now discuss the matching techniques in detail.
4Determination of Equivalence Classes The list of words to be matched isTrst pruned using the areas and aspect ratios of the word images.The pruned list of words is then matched using a matching algorithm. 4.1Pruning
It is assumed that
(1)
where is the area of the template and is the area of the word to be matched.Typical values of used in the experiments range between1.2and1.3.A sim-ilarTltering step is performed using aspect ratios(ie.the width/height ratio).It is assumed that
(2)
The value of used in the experiments range between1.4 and1.7.In both the above equations,the exact factors are not important but it should not be so large so that valid words are omitted,nor so small so that too many words are passed onto the matching phase.
4.2Matching
The template is then matched against the image of each word in the pruned list.The matching function must satisfy two criteria:
1.It must produce a low match error for words which
are similar to the template.
2.It must produce a high match error for words which
are dissimilar.
Two matching algorithms have been tried.TheTrst al-gorithm-Euclidean Distance Mapping(EDM)-assumes that no distortions have occured except for relative transla-tion and is fast.This algorithm usually ranks the matched words in the correct order(i.e.valid wordsTrst,followed by invalid words)when the variations in words is not too large.Although,it returns the lowest errors for words which are similar to the template,it also returns low er-rors for words which are dissimilar to the template.The second algorithm[9],referred to as SLH here,assumes an afTne transformation between the words.It thus compen-sates for some of the variations in the words.This algo-rithm not only ranks the words in the correct order for all examples tried so far,it also seems to be able to better dis-criminate between valid words and invalid words.As cur-rently implemented the SLH algorithm is much slower than the EDM algorithm(we expect to be able to speed it up).5Using Euclidean Distance Mapping for Matching
This approach is similar to that used by[3]to match
machine generated fonts.A brief description of the method
follows(more details are available from[7]).
Consider two images to be matched.There are three
steps in the matching:
1.First the images are roughly aligned.In the vertical
direction,this is done by aligning the baselines of the two images.In the horizontal direction,the images
are aligned by making their left hand sides coincide.
The alignment is,therefore,expected to be accurate in
the vertical direction and not as good in the horizontal direction.This is borne out in practice.
2.Next the XOR image is computed.This is done by
XOR’ing corresponding pixels(see Figure1).
3.An Euclidean distance mapping[2]is computed from
the XOR image by assigning to each white pixel in the
image,its minimum distance to a black pixel.Thus a white pixel inside a blob is assigned a larger distance
than an isolated white pixel.An error measure
can now be computed by adding up the distance mea-sures for each pixel.
4.Although the approximate translation has been com-
puted using step1,this may not be accurate and may
need to beTne-tuned.Thus steps(2)and(3)are re-
peated while sampling the translation space in both x and y.A minimum error measure is com-
puted over all the translation samples.
6SLH Algorithm for Matching
The EDM algorithm does not discriminate well between
good and bad matches.In addition,it fails when there is signiTcant distortion in the words.This happens with the
writing of Erasmus Hudson(Figure2).Thus a matching
algorithm which models some of the variation is needed.
A second matching algorithm(SLH),which models the
distortion as an afTne transformations,was therefore tried
(note that it is expected that the real variation is probably much more complex).An afTne transform is a linear trans-
formation between coordinate systems.In two dimensions, it is described by
(3) where is a2-D vector describing the translation,is
a2by2matrix which captures the deformation,and are the coordinates of corresponding points in the two images between which the afTne transformation must be
recovered.An afTne transform allows for the following
Figure2:Part of a page from the collected papers of the Hudson family
deformations-scaling in both directions,shear in both di-rections and rotation.
The algorithm chosen here is one proposed by Scott and Longuet-Higgins[9](see[6]).The algorithm recovers the correspondence between two sets of points I and J under an afTne transform.This algorithm will now be described.
Two sets of points I and J are created as follows.Every white pixel in theTrst image is a member of the set I.Simi-larly,every white pixel in the second image is a member of set J.First,the centroids of the point sets are computed and the origins of the coordinate systems is set at the centroid. An adjacency matrix G is then computed.The entries are Gaussian weighted distances between a point i in set I and a point j in set J.Each entry is given by
(4) where is the Euclidean distance between i and j.The matrix G is then diagonalized using singular value decom-position(SVD)to give
(5) where D is a diagonal matrix and T and P are orthogonal matrices.The diagonal entries in D are replaced by1’s to give an m by n matrix E.The pairing matrix P
(6) indicates the strength of the attraction between points i and j.Thus a correspondence between two points i and j is posited only if the entry is the greatest element in row i and the greatest element in column j.Intuitively P
is the matrix which correlates best with the G matrix in
the sense of maximizing the trace of.The transfor-mation can then be computed using the recovered corre-
spondence.Scott and Longuet-Higgins showed that if is chosen large enough,the method would compute the cor-respondence correctly for translations,scale changes(i.e. expansions,contractions)and shears.Here,as in inten-sity based algorithms large values of sigma are useful in recovering large translations.However,the method cannot be shown to compute the correct correspondence if a rota-tion is involved.In practice,small rotations can be handled most of the time.
Note that some points will have no correspondence i.e
what the algorithm returns is a one to one correspondence
between some subset of I and some subset of J.
Given the(above)correspondence between point sets
I and J,the afTne transform can be determined by
minimizing the following least mean squares criterion:
(7)
where are the(x,y)coordinates of point and respectively.
The values are then plugged back into the above equa-tion to compute the error.The error is an
estimate of how dissimilar two words are and the words can,therefore,be ranked according to it.
It will be assumed that the variation for valid words is not too large.This implies that if and are con-siderably different from1,the word is probably not a valid match.
Note:The SLH algorithm assumes that pruning on the basis of the area and aspect ratio thresholds is performed. 7Experiments
The two matching techniques were tested on two handwritten pages,each written by a differ-ent writer.TheTrst page can be obtained from the DIMUND document server on the internet https://www.wendangku.net/doc/fe983486.html,/resources/database/hand-writing.database.html This page will be referred to as the Senior document.The handwriting on this page is fairly neat(see[7]for a picture).The second page is from an actual archival collection-the Hudson collection from the library of the University of Massachusetts(part of the page is shown in Figure(2).This page is part of a letter written by James S.Gibbons to Erasmus Darwin Hudson. The handwriting on this page is difTcult to read and the indexing technique helped in deciphering some of the words.
The experiments will show examples of how the match-ing techniques work on a few words.For more examples of the EDM technique see[7].For more examples using the SLH technique and comparisons with the EDM technique see[6].In general,the EDM method ranks most words in the Senior document correctly but ranks some words in the Hudson document incorrectly.The SLH technique per-forms well on both documents.
Both pages were segmented into words(see[7]for de-tails)The algorithm was then run on the segmented words. In the followingTgures,theTrst word shown is the tem-plate.After the template,the other words are ranked ac-cording to the match error.Note that only theTrst few results of the matching are shown although the template has been matched with every word on the page.The area threshold was chosen to be1.2and the aspect ra-tio threshold was chosen as1.4.The translation values were sampled to within pixels in the X direction and pixel in the y direction.Experimentally,this gave the best results.
7.1Results using Euclidean Distance Mapping
The Euclidean Distance Mapping algorithm works rea-sonably well on the Senior document.An example is shown below.
In Figure(3),the template is the word“Lloyd”.TheTg-ure shows that the four other instances of“Lloyd”present in the document are ranked before any of the other words. As Table(1)shows,the match errors for other instances of“Lloyd”is less than that for any other word.In the ta-ble,theTrst column is the Token number(this is needed for identiTcation purposes),the second column is a tran-scription of the word,the third column shows the area in pixels,the fourth gives the match error and the last two columns specify the translation in the x and y directions re-spectively.Note the signiTcant change in area of the words.
Figure3:Ranked matches for template“Lloyd”using the EDM algorithm(the rankings are ordered from left to right and from top to bottom).
Token Word Area Xshift Yshift 105Lloyd13600.00000
70Lloyd12240.17400
165Lloyd12300.175-20
197Lloyd14000.19440
239Lloyd13200.197-30
21Maybe11470.199-10 180along11560.20010
215party12090.20210
245spurt11700.205-10 121dreary14350.20630 Table1:Rankings and match Errors for template“Lloyd”.
The performance on other words in the Senior docu-ment is comparable(for other examples see[7]).This is because the page is written fairly neatly.The performance of the method is expected to correlate with the quality of the handwriting.This was veriTed by running experiments on a page from the Hudson collection(Figure2).The handwriting in the Hudson collection is difTcult to read even for humans looking at grey-level images at300dpi The writing shows wide variations in size-for example, the area of the word“to”varies by as much as100%! However,this large a variation is not expected to occur and
is not seen when the words are larger.Since humans have difTculty reading this material,we do not expect that the
method will perform very well on this document.
The Euclidean Distance Mapping technique fails for the template“Standard”in the Hudson document(see Figure
(4)).The failure occurs because the two instances of“Stan-dard”are written differently.The template“Standard”has a gap between the“t”and the“a”.This gap is not present in
the second example of“Standard”(this is more clearly vis-ible in Figure(6).A technique to model some distortions is,therefore,necessary.
Figure4:Rankings for template“Standard”using the
EDM algorithm(the rankings are ordered from left to right and from top to bottom).
7.2Experiments Using the SLH Algorithm
The SLH algorithm handles afTne distortions and is,
therefore more powerful then the EDM algorithm.Since the current version of the SLH algorithm is slow,the initial
matches were pruned using the EDM algorithm and then the SLH algorithm run on the pruned subset.
Experiments were performed using both the Senior doc-
ument and the Hudson documents.A few examples are shown here(for more details see[6]).For the Senior doc-uments the same pruning ratios were chosen as before.
To account for the large variations in the Hudson papers, the area threshold wasTxed at1.3and the aspect ratio threshold at1.7.The value of depends on the expected
translation.Since it is small,.A lower value of yielded poorer results.
The matches for the template“Lloyd”are shown in Ta-
ble(2).The succesive columns of the table,tabulate the Token Number,the transcription of the word,the area of the word image,the number of corresponding points recov-
ered by the SLH algorithm,the match error using the SLH algorithm and the afTne transform.The entries
are ranked according to the match error.If either of or is less than0.8or greater than1/0.8,that word is eliminated from the rankings.A comparison with Table(1)shows that the rankings change.This is not only true of the invalid words(for example the sixth entry in Table(1)is“Maybe”while the sixth entry in Table(2)is “lawyer”but is also true of the“Lloyd”’s.Both tables rank instances of“Lloyd”ahead of other words.The technique also shows a much greater discrimination in match error -the match error for“lawyer”is almost double the match error for theTfth“Lloyd”.
In English,theTrst letter in a word is capitalized when the word begins a sentence and not otherwise(unless it is a proper noun).Thus it is desirable that the technique be relatively insensitive to this capitalization.Figure(5) shows an example of this.The word“minister”is the high-est ranked word obtained for the template“Minister”de-spite the fact that“minister”begins with a lower case letter while“Minister”starts with an uppercase letter.
Figure5:Rankings for template“Minister”for the SLH algorithm(the rankings are ordered from left to right and from top to bottom).
The method was also run on the Hudson document(Fig-ure(2))and it ranked most of the words correctly on this document.As an example,we look at the word“Standard”on which the EDM method did not rank correctly.The SLH method produces the correct ranking inspite of the signiTcant distortions in the word(see Figure(6)).
8Conclusion
The work clearly shows that the idea of indexing a cor-pus of written words in a single hand is feasible.Two dif-ferent matching algorithms were tried.TheTrst based on Euclidean Distance Matching was a simple matching algo-rithm which accounted for only translational shifts.This algorithm was able to match most of the words when the handwriting was good.The performance degrades with the quality of the handwriting,although surprisingly it does reasonably well on poor quality handwriting.The second algorithm models the transformation between words as an afTne transform.This technique has so far been able to handle most of the matching problems associated with the Euclidean distance technique[6].
Some personal digital assistants like the Newton allow handwritten entries using a(digital)pen.This technique can be used to retrieve previously stored material by using the word matching techniques suggested here.
9Acknowledgements
The original idea of using word spotting to index hand-written documents was suggested by Bruce Croft.We
Token Word Area CP A T
105Lloyd13682330.00 1.000.000.00
0.00 1.000.00
197Lloyd1400199 1.3020.96-0.04 1.58
0.01 1.040.14
70Lloyd1224176 1.3560.940.09-1.02
0.030.92-1.38
165Lloyd1230189 1.631 1.030.05-0.43
-0.010.87-2.60 239Lloyd1320203 1.7950.99-0.05 1.44
0.03 1.07 2.21
157lawyer1518185 3.3930.96-0.03 1.89
0.05 1.110.03
240Selwyn1564188 3.6730.940.06-4.23
0.05 1.05-0.75
91thought1178181 3.9730.970.03 2.33
-0.01 1.08 2.91
Table2:Rankings and Match Errors for template“Lloyd”Using SLH Algorithm.
Figure6:Rankings for template“Standard”for the SLH algorithm(the rankings are ordered from left to right and from top to bottom).
would like to thank Gail Giroux and the University of Mas-sachusetts Library for the scanned page from the Hudson collection.We also wish to thank Bob Heller and Jonathan Lim for systems help.
References
[1]M.Bokser.Omnidocument technologies.Proceedings
IEEE,80(7):1066–1078,1992.
[2]Per-Erik Danielsson.Euclidean distance https://www.wendangku.net/doc/fe983486.html,-
puter Graphics and Image Processing,14:227–248,1980.
[3]Paul Filiski and Jonathan J.Hull.Keyword selection from
word recognition results using deTnitional overlap.In Third Annual Symposium on Document Analysis and Information Retrieval,UNLV,Las Vegas,pages151–160,1994.
[4]G.J.F.Jones,J.T.Foote,K.Sparck Jones,and S.J.Young.
Video mail retrieval:The effect of word spotting accuracy on precision.In International Conference on Acoustics, Speech and Signal Processing,volume1,pages309–316, 1995.
[5]Siamak Khoubyari and Jonathan J.Hull.Keyword location
in noisy document image.In Second Annual Symposium on Document Analysis and Information Retrieval,UNLV,Las Vegas,pages217–231,1993.
[6]R.Manmatha,Chengfeng Han,and E.M.Riseman.Word
spotting:A new approach to indexing handwriting.Tech-nical Report CS-UM-95-105,Computer Science Dept,Uni-versity of Massachusetts at Amherst,MA,1995.
[7]R.Manmatha,Chengfeng Han,E.M.Riseman,and W.B.
Croft.Indexing handwriting using word matching.In Dig-ital Libraries’96:1st ACM International Conference on Digital Libraries,1996.
[8]S.Mori,C.Y.Suen,and K.Yamamoto.Historical review
of ocr research and development.Proceedings of the IEEE, 80(7):1029–1058,1992.
[9]G.L.Scott and H.C.Longuet-Higgins.An algorithm for
associating the features of two patterns.Proc.Royal Society of London B,B244:21–26,1991.
[10]H.R.Turtle and W.B.Croft.A comparison of text retrieval
https://www.wendangku.net/doc/fe983486.html,puter Journal,35(3):279–290,1992.
市政工程等级划分标准[1]
市政工程类别划发标准及有关规定 （一）市政工程类别划分标准 一类工程： 1.道路工程： (1)六条机动车道及以上的高级路面或次高级路面； (2)四条机动车道及以上的并有绿化隔离带的非机动车道的高级或次高级路面。 2.机场跑道、停机坪。 3.桥涵工程： (1)二层或桥面最高高度16m及以上立交桥及路面； (2)单跨跨径Ｌo≥40m工程； (3)多跨总长度Ｌ≥100m工程。 4.最大管径DN≥1200mm及总长度Ｌ≥1200m的排水工程。 5.最大管么DN≥1000mm及总长度Ｌ≥1000m的给水工程。 6.机械项管工程及箱体最大宽度≥6m的渠箱沉井工程。 7.最大管径DN≥630mm的燃气工程。 8.高3m以上厚1m及以上的防洪墙工程。 9.截面净宽度Bo≥90m的隧道工程。 二类工程： 1.道路工程：四条机动车道及以上的高级路面或次高级路面。 2.桥涵工程： (1) 一层或桥面最高高度小于16m的立交桥及路面。 (2)单跨跨径Lo≥20m; (3)多跨长度Ｌ≥70m的桥涵工程。
3.最大管径DN≥1000mm及总长度L≥1000m的排水工程。 4.最大管径DN≥600mm及总长度Ｌ≥800m的给水工程。 5.最大管径DN≥325mm的燃气工程。 6.具有明渠、暗渠及截洪沟单独排洪工程。 7.截面净宽度Bo≥7m的隧道工程。 8.城市道路照明工程。 9.高3m以下，厚1m以下的防洪墙。 三类工程： 1.二条机动车及以上的混凝土、沥青混凝土路面的道路工程。 2.桥涵工程： (1)单跨跨径Lo≥10m桥涵工程； (2)多跨总长度L≥30m的桥涵工程。 3.最大管径DN≥500mm及总长度L≥500m的排水工程。 4.最大管径DN≥300mm及总长度L≥400m的给水工程。 5.最大管径DN<325mm的燃气工程。 6.具有抛石石笼砌护底护脚及其相应配套的土石方附属工程等的单独护岸护坡工程。 7.截面净宽度Bo<7m的隧道工程。 8.截面净空面积So≥4m2的砌筑沟渠。 四类工程： 1.道路工程： (1)二条机动车道（不含快速路）以下工程； (2)单独人行道。 2.桥涵工程：
香港大学研究生科目排行榜
香港各大学2001年度排名榜 最佳会计/专业会计(BEST ACCOUNTING/ PROFESSIONAL ACCOUNTANCY) 香港理工大学(The Hong Kong Polytechnic University) 最佳人类学/人文科学(BEST ANTHROPOLOGY/ HUMANITIES) 香港中文大学(The Chinese University of Hong Kong ) 最佳建筑学(BEST ARCHITECTURAL STUDIES ) 香港大学(The University of Hong Kong ) 最佳文学(BEST ARTS ) 香港中文大学(The Chinese University of Hong Kong ) 最佳生物学/应用生物(BEST BIOLOGY/ APPLIED BIOLOGY ) 香港科技大学(The Hong Kong University of Science and Technology ) 最佳生物化学(BEST BIOCHEMISTRY ) 香港中文大学(The Chinese University of Hong Kong ) 最佳工商管理学(BEST BUSINESS ADMINISTRATION ) 香港中文大学(The Chinese University of Hong Kong ) 最佳屋宇设备工程学(BEST BUILDING SERVICES ENGINEERING ) 香港理工大学(The Hong Kong Polytechnic University ) 最佳化学/应用化学/化学科技(BEST CHEMISTRY/ APPLIED CHEMISTRY/ CHEMICAL TECHNOLOGY ) 香港大学(The University of Hong Kong ) 最佳中国语文及文学(文学学士) (BEST CHINESE LANGUAGE AND LITERATURE (BA) ) 香港中文大学(The Chinese University of Hong Kong) 最佳教育(中文) (BEST EDUCATION) (CHINESE) 香港中文大学( The Chinese University of Hong Kong) 最佳教育(英文) (BEST EDUCATION) (ENGLISH) 香港大学(The University of Hong Kong ) 最佳中医(BEST CHINESE MEDICINE) 香港中文大学(The Chinese University of Hong Kong) 最佳土木工程学(BEST CIVIL ENGINEERING)
路面等级与分类
(三)路面等级与分类 1.路面等级 路面等级按面层材料的组成、结构强度、路面所能承担的交通任务与使用的品质划分为高级路面、次高级路面、中级路面与低级路面等四个等级。 2.路面类型 (1)路面基层的类型。按照现行规范,基层(包括底基层)可分为无机结合料稳定类与粒料类。无机结合料稳定类有:水泥稳定土、石灰稳定土、石灰工业废渣稳定土及综合稳定土;粒料类分级配型与嵌锁型,前者有级配碎石(砾石),后者有填隙碎石等。 ①水泥稳定土基层。在粉碎的或原来松散的土中,掺人足量的水泥与水,经拌与得到的混合料在压实养生后,当其抗压强度符合规定要求时,称为水泥稳定土。可适用于各种交通类别的基层与底基层,但水泥土不应用作高级沥青路面的基层,只能作底基层。在高速公路与一级公路的水泥混凝土面板下,水泥土也不应用作基层。 ②石灰稳定土基层。在粉碎或原来松散的土中掺入足量的石灰与水,经拌与、压实及养生得到的混合料,当其抗压强度符合规定要求时,称为石灰稳定土。适用于各级公路路面的底基层,可作二级与二级以下的公路的基层,但不应用作高级路面的基层。 ③石灰工业废渣稳定土基层。一定数量的石灰与粉煤灰或石灰与煤渣与其她集料相配合,加入适量的水,经拌与、压实及养生后得到的混合料,当其抗压强度符合规定要求时,称为石灰工业废渣稳定土,简称石灰工业废渣。适用于各级公路的基层与底基层,但其中的二灰土不应用作高级沥青路面及高速公路与一级公路上水泥混凝土路面的基层。 ④级配碎(砾)石基层。由各种大小不同粒径碎(砾)石组成的混合料,当其颗粒组成符合技术规范的密实级配的要求时,称其为级配碎(砾)石。级配碎石可用于各级公路的基层与底基层,可用作较薄沥青面层与半刚性基层之间的中间层。级配砾石可用于二级与二级以下公路的基层及各级公路的底基层。 ⑤填隙碎石基层。用单一尺寸的粗碎石作主骨料,形成嵌锁作用,用石屑填满碎石间的空隙,增加密实度与稳定性,这种结构称为填隙碎石。可用于各级公路的底基层与二级以下公路的基层。 (2)路面面层类型。根据路面的力学特性,可把路面分为沥青路面、水泥混凝土路面与其她类型路面。
2015香港大学内地招生简章
2015港校招生政策最新解读 招生类型 港校内地招生分为两大类：一类港校是通过高考统一招生，包括：香港城市大学、香港中文大学，这类学校的报考和内地众多大学的报考流程一样，考生只需要在高考后上网填报志愿即可，高考前并不需要再另外填写申请表。 另一类港校则是通过自主招生录取，包括：香港科技大学，岭南大学，香港大学，香港公开大学，香港教育学院，香港浸会大学，香港理工大学，香港树仁大学，香港演艺学院，香港珠海学院。这类学校要求，申请人除了参加高考以外，高考前，考生还需要在学校规定的时间段内，向学校提出额外的报名申请。 报考香港城市大学及香港中文大学不需要参加面试环节，而除了这两所港校，其他大多数港校都将设立面试环节，且从各个方面考察学生的综合素质，难度不小，建议考生提前准备。 招生计划 部分港校已明确提出招生名额要求，其中香港科技大学以择优方式录取，并非采取计划招生，各省市招生名额不设上下限。2015年招生名额约150人； 2015 年香港城市大学在31 个省市区的本科招生总数为208 名。分省分学院招生计划以及相关说明将分别由高招办或考试院统一向社会公布； 香港大学2015年度内地本科生入学计划预计招收不超过300名学生，各省市不设名额上限，择优录取； 香港理工大学2015-16年度在内地的招生名额约为290名。 香港高校招收内地本科生始于1998年。目前，香港招收内地自费本科生（全日制）的院校有12所：香港大学、香港中文大学、香港理工大学、香港科技大学、香港城市大学、香港浸会大学、岭南大学、香港教育学院、香港公开大学、香港演艺学院、香港树仁大学、珠海学院，招生流程分为两种：港校自主招生录取流程与统招提前批次录取流程。 港校自主招生录取流程 包括香港大学、香港理工大学、香港科技大学在内的多所港校，采用港校自主招生方式，面向内地录取本科学生。港校自主招生过程独立于统招，不会影响统招志愿的填报与录取。 第一步报考这些院校学生，须到相应港校官方网站进行在线报名，并缴纳申请费。报名通常截止于六月初，有意报考港校的考生和家长一定注意不要错过报名时间。 第二步此类港校内地招生需参考高考成绩，因此报名之后学生需按时参加内地高考，然后集中准备港校面试，静待面试通知。在得知高考成绩之后，报考香港理工大学和香港科技大学的考生，须登入港校网上入学申请系统，自行输入高考成绩。 第三步多数港校录取过程设有面试环节，考察考生的综合素质。各大港校将于六月中、
公路等级划分标准
公路等级划分标准 公路是连接各城市、城市和乡村、乡村和厂矿地区的道路。 根据使用任务、功能和适应的交通量分为高速公路、一级公路、二级公路、三级公路、四级公路五个等级。 高速公路是具有特别重要的政治经济意义的公路，有四个或四个以上车道，并设有中央分隔带、全部立体交叉并具有完善的交通安全设施与管理设施、服务设施，专供汽车分向、分车道行驶并全部控制出入的多车道干线专用公路。能适应年平均日交通量（ADT）25000辆以上。四车道高速公路一般能适应按各种汽车折合成小客车的远景设计年限年平均昼夜交通量（ADT）为2500～55000辆；六车道高速公路一般能适应按各种汽车折合小客车的远景设计年限年平均昼夜交通量为45000～80000辆；八车道高速公路一般能适应按各种汽车折合成人客车的远景设计年限年60000～100000辆。 其它公路为除高速公路以外的干线公路、集散公路、地方公路，分四个等级。 一级公路是连接重要政治经济文化中心、部分立交的公路，为供汽车分向、分车道行驶并根据需要控制出入的多车道公路，四车道一级公路一般能适应按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为15000～30000辆；六车道一级公路一般能适应按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为25000～55000辆。 二级公路是连接政治、经济中心或大工矿区的干线公路、或运输繁忙的城郊公路，为供汽车行驶的双车道公路，一般能适应按各种车辆折合成中型载重汽车的远景设计年限年平均昼夜交通量为3000～7500辆（或按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为7500～15000辆）。 三级公路是沟通县或县以上城市的支线公路，为主要供汽车行驶的双车道公路，一般能适应按各种车辆折合成中型载重汽车的远景设计年限年平均昼夜交通量为1000～4000辆（或按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为2000～6000辆）。 四级公路是沟通县或镇、乡的支线公路，为主要供汽车行驶的双车道或单车道公路，一般能适应按各种车辆折合成中型载重汽车的远景设计年限年平均昼夜交通量为：双车道1500辆以下（或按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为2000辆以下）；单车道200辆以下（或按各种汽车折合成小客车的远景设计年限年平均昼夜交通量为400辆以下）。 二、公路根据在政治、经济、国防上的重要意义和使用性质划分为5个行政等级 国家公路（国道）指具有全国性政治、经济意义的主要干线公路，包括重要的国际公路、国防公路，连接首都与各省、自治区、直辖市首府的公路，连接各大经济中心、港站枢纽、商品生产基地和战略要地的干线公路。 省公路（省道）指具有全省（自治区、直辖市）政治、经济意义，连接各地市和重要地区以及不属于国道的干线公路。 县公路（县道）指具有全县（县级市）政治、经济意义，连接县城和县内主要乡（镇）、主要商品生产和集散地的公路，以及不属于国道、省道的县际间公路。 乡公路（乡道）指主要为乡（镇）村经济、文化、行政服务的公路，以及不属于县道以上公路的乡与乡之间及乡与外部联络的公路。 专用公路指专供或主要供厂矿、林区、农场、油田、旅游区、军事要地等与外部联系的公路。 三、路面等级按面层类型分高级、次高级、中级和低级 高级沥青混凝土路面或水泥混凝土路面。 次高级沥青贯入或路面式沥青碎石路面。 中级沙石路面。
土木工程项目管理a第次
土木工程项目管理A 第一次作业 一、单项选择题(只有一个选项正确，共10道小题) 1.业主为了降低工程实施的风险，找了保险公司进行投保，这种行为属于。 (A)风险减轻 (B)风险自担 (C)风险转移 (D)风险共担 正确答案：C 解答参考： 2.（）是项目取得成功的保证 (A)业主 (B)政府 (C)项目管理 (D)项目计划 正确答案：C 解答参考：
3.下列有关工程项目目标的表述中不正确的是（） (A)工程项目管理实际上就是一种目标管理 (B)工程项目目标就是实施一个工程项目所要达到的预期结果 (C)工程项目目标必须明确、可行、具体和可以度量 (D)工程项目目标不一定在投资方与业主、承包商之间达成一致 正确答案：D 解答参考： 4.下列关于项目范围管理的表述中，错误的一项是()。 (A)工程项目范围管理是指确保项目完成全部规定要做的工作，而且仅仅完成规定要做的工作，从而成功的达到项目目标的管理过程，即，在满足工程项目使用功能的条件下，对项目应该包括哪些具体的工作进行定义和控制 (B)工程项目范围管理的内容包括工程项目范围定义、项目范围确认和范围的变更控制 (C)范围一词应解释为包括下述两方面的含义：一是工程项目将要包括的性质和使用功能；二是实施并完成该工程项目而必须做的具体工作 (D)范围管理在工程项目建设周期的各个阶段的内容相同 正确答案：D 解答参考： 5.关于对范围确认的表述，错误的一项是()。 (A)范围确认是项目业主正式接收项目工作成果的过程
(B)此过程要求对项目在执行过程中完成的各项工作进行及时的检查，保证正确地、满意地完成合同规定的全部工作 (C)如果项目提前终止，范围确认过程也应确定和正式记录项目完成的水平和程度 (D)范围确认与质量控制相等同 正确答案：D 解答参考： 6.矩阵式组织结构的缺点的表述中，错误的一项为( )。 (A)项目管理权力平衡困难。矩阵式组织结构中项目管理的权力需要在项目经理与职能部门之间平衡，这种平衡在实际工作中是不易实现的 (B)信息回路比较复杂。在这种模式下，信息回路比较多，即要在项目团队中进行，还要在相应的部门中进行，必要时在部门之间还要进行，所以易出现交流、沟通不够的问题 (C)项目成员处于多头领导状态。项目成员正常情况下至少要接受二个方向的领导，即项目经理和所在部门的负责人，容易造成指令矛盾，行动无所适从的问题 (D)项目组织成为一个相对封闭的组织，公司的管理与对策在项目管理组织中贯彻可能遇到阻碍（项目式的缺点) 正确答案：D 解答参考： 7.根据项目的总目标和阶段性目标，将项目的最终成果和阶段性成果进行分解，列出达到这些目标所需的硬件和软件，它实际上是对子项目或项目的组成部分进一步分解形成的结构图表，其主要技术是按工程内容进行项目分解，这一过程不包括()。
道路等级划分
道路等级划分 道路等级划分道路指供各种车辆（无轨）和行人等通行的工程设施。道路包含众多种类，性质功能等均有不同，因此无法用一个唯一标准对所有道路进行等级划分。为此各国现行做法一般都是先划分道路种类，后针对各类道路的技术标准划分等级。 我国按照道路使用特点，可分为城市道路.公路.厂矿道路.林区道路和乡村道路。除对公路和城市道路有准确的等级划分标准外，对林区道路.厂矿道路和乡村道路一般不再划分等级。 城市道路是指在城市范围内具有一定技术条件和设施的道路。根据道路在城市道路系统中的地位.作用.交通功能以及对沿线建筑物的服务功能，我国目前将城市道路分为四类：快速路.主干路.次干路及支路。其中快速路在特大城市或大城市中设置，是用中央分隔带将上，下行车辆分开，供汽车专用的快速干路，主要联系市区各主要地区.市区和主要的近郊区.卫星城镇.联系主要的对外出路，负担城市主要客.货运交通，有较高车速和大的通行能力。主干路是城市道路网的骨架，联系城市的主要工业区.住宅区.港口.机场和车站等额货运中心，承担着城市主要交通任务的交通干道。主干路沿线两侧不宜修建过多的行人和车辆入口，否则会降低车速。次干路为市区内普通的交通干路，配合主干路组成城市干道网，起联系各部分和集散作用，分担主干路的交通负荷。次干路兼有服务功能，允许两侧布置吸引人流的公共建筑，
并应设停车场。支路是次干路与街坊路的连接线，为解决局部地区的交通而设置，以服务功能为主。部分主要支路可设公共交通线路或自行车专用道，支路上不宜有过境交通。 根据国家《城市规划定额指标暂行规定》的有关规定，道路还可划分为四级，如表12-2所示：表12-2 道路四级划分表项目级别设计车速(km/h)双向机动车道数（条）机动车道宽度(m)道路总宽(m)分隔带设置一级６0～80 >=43.7540～70 必须设二级40～60 >=43.530～60 应设三级30～40 >=23.520～40 可设四有30 >=23.516～30 不设公路是连接各城市.城市和乡村.乡村和厂矿地区的道路。根据交通量.公路使用任务和性质，将公路分为以下五个等级：高速公路：是具是有特别重要的政治经济意义的公路，有四个或四个以上车道，并设有中央分隔带.全部立体交叉并具有完善的交通安全设施与管理设施.服务设施，全部控制出入，专供汽车高速行驶的专用公路。能适应年平均日交通量（ADT）25000辆以上。 一级公路：是连接重要政治经济文化中心.部分立交的公路，一般能适应ADT=10000～25000辆。 二级公路：是连接政治.经济中心或大工矿区的干线公路.或运输繁忙的城郊公路，能适应ADT=2000～10000辆。 三级公路：是沟通县或县以上城市的支线公路，能适应 ADT=200～2000辆。
土木工程项目管理模拟试题
北京科技大学远程教育学院 土木工程项目管理 试题（模拟1） 学号： 学习中心名称： 专业： 层次： 姓名： 一、多项选择题（共10题，每题2分。每题的备选项中，有2个或2个以上 符合题意，至少有一个错项。错选，本题不得分；少选，所选的每个选项得0.5 分）： （2分）1、建设项目管理的内涵是：自项目开始至项目完成，通过项目策划和项目控制使项目（ ）的三大目标得以实现。 A ．费用 B ．协调 C ．质量 D ．进度 E ．合同 （2分）2．投资的计划值和实际值是相对的，相对于工程合同价，（ ）可作为投资的计划值。 A ．工程概算 B ．工程预算 C ．工程款支付 D ．工程决算 E ．工程中标价 （2分）3、施工成本控制可分为（ ）等控制内容和工作。 A ．程序控制 B ．事先控制 C ．过程控制 D ．事后控制 E ．全员控制 装 订 线 内 不 得 答 题 自 觉 遵 守 考 试 规 则，诚 信 考 试，绝 不 作 弊
（2分）4、某混凝土工程某月计划工程量为110m3，计划成本为320元/ m3，月底检查时承包商实际完成工程量为100 m3，实际成本为300元/m3，则下列关于该工程施工成本偏差和进度偏差（用成本表示）的表述，正确的是（） A．成本节约2000元B．成本超支2000元C．成本节约5200元D．工期提前3200元E．工期拖后3200元 （2分）5、在各种计划方法中，（）的工作进度线与时间坐标相对应。 A．形象进度计划B．横道图计划C．双代号网络计划D．单代号答接网络计划E．双代号时标网络计划 （2分）6、建设工程招标应具备的条件是（） A．招标人已经依法成立 B．初步设计及概算应当履行审批手续，已经批准 C．招标范围、方式和组织形式等应当履行核准手续的，已经核准 D．有相应资金或资金来源已经落实，且有招标所需设计图纸及技术资料 E．发布了招标公告 （2分）7、竣工验收合格即办理移交。移交表示（） A．发包人认可并接收工程，承包人工程施工任务的完结 B．工程所有权的转让C．发包人工程照管责任的结束和承包人工程照管责任的开始D．保修责任的开始E．合同规定的工程款支付条款有效 （2分）8、建设工程索赔的起因是（）
香港大学研究生科目排行榜
最佳会计/专业会计(BEST ACCOUNTING/ PROFESSIONAL ACCOUNTANCY) 香港理工大学(The Hong Kong Polytechnic University) 最佳人类学/人文科学(BEST ANTHROPOLOGY/ HUMANITIES) 香港中文大学(The Chinese University of Hong Kong ) 最佳建筑学(BEST ARCHITECTURAL STUDIES ) 香港大学(The University of Hong Kong ) 最佳文学(BEST ARTS ) 香港中文大学(The Chinese University of Hong Kong ) 最佳生物学/应用生物(BEST BIOLOGY/ APPLIED BIOLOGY ) 香港科技大学(The Hong Kong University of Science and Technology ) 最佳生物化学(BEST BIOCHEMISTRY ) 香港中文大学(The Chinese University of Hong Kong ) 最佳工商管理学(BEST BUSINESS ADMINISTRATION ) 香港中文大学(The Chinese University of Hong Kong ) 最佳屋宇设备工程学(BEST BUILDING SERVICES ENGINEERING ) 香港理工大学(The Hong Kong Polytechnic University ) 最佳化学/应用化学/化学科技(BEST CHEMISTRY/ APPLIED CHEMISTRY/ CHEMICAL TECHNOLOGY ) 香港大学(The University of Hong Kong ) 最佳中国语文及文学(文学学士) (BEST CHINESE LANGUAGE AND LITERATURE (BA) ) 香港中文大学(The Chinese University of Hong Kong) 最佳教育(中文) (BEST EDUCATION) (CHINESE) 香港中文大学( The Chinese University of Hong Kong) 最佳教育(英文) (BEST EDUCATION) (ENGLISH) 香港大学(The University of Hong Kong ) 最佳中医(BEST CHINESE MEDICINE) 香港中文大学(The Chinese University of Hong Kong) 最佳土木工程学(BEST CIVIL ENGINEERING) 香港大学(The University of Hong Kong)
2015-CAFIC_IACIS_Call for Papers——香港理工大学
The 21st Intern Association for Inte The 11th Biennual Intern Intercul Conference Theme: “Culture, C Conference Time: July 15-18Conference Venue: The Hong K Conference Goals: ? To provide scholars, educato opportunities to interact, netw intercultural communication ? To synthesize research perspe integrated approaches to com ? To advance the methodology findings to facilitate understa ? To foster global intercultural and other stakeholders world communication issues. Topic areas are broadly defined Advertising and marketing ac Bilingual and multilingual co Bilingual education Comparative literature Conflict, mediation and nego Cultures Corporate culture and manag Crisis/risk communication Cross-cultural adaptation Cross-cultural interaction Culture and diplomacy Digital communication acros Ethnic studies Gender issues and communic Glocalization Group/Organizational comm Health communication Intercultural communication Intercultural communication nternational Conference of the Intern or Intercultural Communication Stud cum International Conference of the China tercultural Communication (CAFIC)Call for Submissions ure, Communication, and Hybridity in an Age of G 8, 2015 ong Kong Polytechnic University, Hong Kong ucators and practitioners from different cultural co t, network and benefit from each other’s research a ation issues; perspectives and foster interdisciplinary scholarly d o complex problems of communication across cult ology for intercultural communication research an derstanding across cultures; ltural sensitivity and involve educators, business p worldwide in the discourse about diversity and tran efined as, but not limited to, the following: ing across cultures ual communication d negotiation across management ion on across cultures munication communication ation competence ation and politics Intercultural and multili Intercultural pragmatics Interethnic communicat Interpersonal communic Language and cultural h Language planning and Media and transcultural Multiple cultures and in Philosophy and human b Psychological communi Public opinions and pub Public relations Rhetorical communicati Second language educat Social media and comm Sustainability and globa Transculturality in globa Translation studies Verbal and nonverbal co International Studies (IAICS) China Association for AFIC) e o f Globalization ” ral communities with arch and expertise related to larly dialogues for developing s cultures; ch and disseminate practical ness professionals, students nd transcultural multilingual education matics unication and relations munication and relations ural hybridity g and policy ltural communication and interculturality man behavior patterns mmunication studies d public policy nication education communication globalization global context bal communication
道路工程验收表格
分项工程质量检验记录 质验表（路）—79 建设单位：苏州市市政建设管理处监理单位：南京旭光建设监理有限公司合同号： 施工单位：苏州盛昌市政工程有限公司编号： 单位(子单位)工程名称道路工程 分部(子分部)工程名称路基分项工程名称6%灰土土方路基检验批数项目经理方华明 序号检验批部位、区段施工单位检查评定结果监理（建设）单位验收结论 1 K1+160-K1+233东侧第1层 2 K1+160-K1+233东侧第2层 3 K1+233-K1+320东侧第1层 4 K1+233-K1+320东侧第2层 5 K1+320-K1+460东侧第1层 6 K1+320-K1+460东侧第2层 7 K1+460-K1+675东侧第1层 8 K1+460-K1+675东侧第2层 9 K1+755-K1+870东侧第1层 10 K1+755-K1+870东侧第2层 11 K1+870-K1+920东侧第1层 12 K1+870-K1+920东侧第2层 13 K1+920-K1+964东侧第1层 14 K1+920-K1+964东侧第2层 15 K1+964-K2+000东侧河塘回 填第1层 16 K1+964-K2+000东侧河塘回 填第2层 17 K2+000-K2+050东侧第1层 18 K2+000-K2+050东侧第2层 施工单位检查评定结果 专业工长（施工员）施工班组长 质量检查员：年月日
专业监理工程师：年月日 分项工程质量检验记录 质验表(路）—79 建设单位：苏州市市政建设管理处监理单位：南京旭光建设监理有限公司合同号： 施工单位：苏州盛昌市政工程有限公司编号： 单位(子单位)工程名称道路工程 分部(子分部)工程名称路基分项工程名称6%灰土土方路基检验批数项目经理方华明 序号检验批部位、区段施工单位检查评定结果监理（建设）单位验收结论 19 K1+160-K1+233西侧第1层 20 K1+160-K1+233西侧第2层 21 K1+233-K1+340西侧第1层 22 K1+233-K1+340西侧第2层 23 K1+340-K1+500西侧第1层 24 K1+500-K1+640西侧第1层 25 K1+340-K1+640西侧第2层 26 K1+720-K1+870西侧第1层 27 K1+870-K2+008西侧第1层 28 K1+720-K1+870西侧第2层 29 K1+870-K2+008西侧第2层 30 K2+008-K2+050西侧河塘回 填第1层 31 K2+008-K2+050西侧河塘回 填第2层 32 华元路南侧东喇叭口回填 第1层 33 华元路南侧东喇叭口回填 第2层 34 华元路南侧东喇叭口回填 第3层 施工单位检查评定结果 专业工长（施工员）施工班组长 质量检查员：年月日
土木工程项目管理A第~次
土木工程项目管理A第 ~次 IMB standardization office【IMB 5AB- IMBK 08- IMB 2C】
土木工程项目管理 A 第一次作业 一、单项选择题(只有一个选项正确，共10道小题) 1.业主为了降低工程实施的风险，找了保险公司进行投保，这种行为属于。 (A)风险减轻 (B)风险自担 (C)风险转移 (D)风险共担 正确答案：C 解答参考： 2.（）是项目取得成功的保证 (A)业主 (B)政府 (C)项目管理 (D)项目计划 正确答案：C 解答参考： 3.下列有关工程项目目标的表述中不正确的是（） (A)工程项目管理实际上就是一种目标管理 (B)工程项目目标就是实施一个工程项目所要达到的预期结果 (C)工程项目目标必须明确、可行、具体和可以度量 (D)工程项目目标不一定在投资方与业主、承包商之间达成一致 正确答案：D 解答参考： 4.下列关于项目范围管理的表述中，错误的一项是()。 (A)工程项目范围管理是指确保项目完成全部规定要做的工作，而且仅仅完成规定要做的工作，从而成功的达到项目目标的管理过程，即，在满足工程项目使用功能的条件下，对项目应该包括哪些具体的工作进行定义和控制
(B)工程项目范围管理的内容包括工程项目范围定义、项目范围确认和范围的变更控制 (C)范围一词应解释为包括下述两方面的含义：一是工程项目将要包括的性质和使用功能；二是实施并完成该工程项目而必须做的具体工作 (D)范围管理在工程项目建设周期的各个阶段的内容相同 正确答案：D 解答参考： 5.关于对范围确认的表述，错误的一项是()。 (A)范围确认是项目业主正式接收项目工作成果的过程 (B)此过程要求对项目在执行过程中完成的各项工作进行及时的检查，保证正确地、满意地完成合同规定的全部工作 (C)如果项目提前终止，范围确认过程也应确定和正式记录项目完成的水平和程度 (D)范围确认与质量控制相等同 正确答案：D 解答参考： 6.矩阵式组织结构的缺点的表述中，错误的一项为( )。 (A)项目管理权力平衡困难。矩阵式组织结构中项目管理的权力需要在项目经理与职能部门之间平衡，这种平衡在实际工作中是不易实现的 (B)信息回路比较复杂。在这种模式下，信息回路比较多，即要在项目团队中进行，还要在相应的部门中进行，必要时在部门之间还要进行，所以易出现交流、沟通不够的问题 (C)项目成员处于多头领导状态。项目成员正常情况下至少要接受二个方向的领导，即项目经理和所在部门的负责人，容易造成指令矛盾，行动无所适从的问题 (D)项目组织成为一个相对封闭的组织，公司的管理与对策在项目管理组织中贯彻可能遇到阻碍（项目式的缺点) 正确答案：D 解答参考： 7.根据项目的总目标和阶段性目标，将项目的最终成果和阶段性成果进行分解，列出达到这些目标所需的硬件和软件，它实际上是对子项目或项目的组成部分进一步分解形成的结构图表，其主要技术是按工程内容进行项目分解，这一过程不包括()。 (A)识别项目的主要组成部分 (B)确定该级别的每一单元是否可以恰当的估算费用和工期。每个单元应该有相同或相近的分解级别，以便实施进度测量 (C)识别每一可交付成果的组成单元 (D)证实分解的正确性
2015年暂停一次科学院院士申报资格的人
中国科学院2015年暂停一次申报资格的人数学物理学部 41 郑伟谋1946 中科院理论物理研究所福建福州03 05 09 11 13 36 段海豹1956 中科院数学与系统科学研究院陕西西安09 11 13 (百人) 46 盛政明1966 上海交通大学浙江嘉兴09 11 13 (杰青) 03 王牧 1962 南京大学江苏扬州09 11 13 (杰青) 化学部 26 陈国南1948 福州大学福建福州09 11 13 03 王柯敏1957 湖南大学湖南宁乡09 11 13 (杰青) 06 帅志刚1962 清华大学江西铅山09 11 13 (杰青) 40 谢作伟1964 香港中文大学浙江苍南09 11 13 (杰青B) 02 王利祥1964 中科院长春应用化学研究所黑龙江 09 11 13 (杰青) 23 杨金龙1966 中国科学技术大学江苏盐城09 11 13 (杰青) 29 岳建民1962 中国科学院上海药物研究所 09 11 13 (杰青) 生命科学和医学学部 63 瞿佳 1955 温州医学院浙江温州09 11 13 43 郭亚军1955 第二军医大学山东莘县09 11 13 (杰青) 03 王岩 1961 中国人民解放军总医院北京 09 11 13 27 陈晔光1964 清华大学江西乐安09 11 13 (杰青) 21 杨晓 1967 军事医学科学院四川 09 11 13 (杰青) (女) 44 徐涛 1970 中科院生物物理研究所湖北宜昌05 09 11 13 (杰青) 地学部 21 陈中原1953 华东师范大学浙江瑾县09 11 13 49 高抒 1956 南京大学浙江杭州03 05 09 11 13 (杰青) 04 史培军1959 北京师范大学陕西靖边09 11 13 (杰青) 09 成秋明1960 中国地质大学(武汉) 山西太谷09 11 13 (杰青) 19 李献华1961 中科院地质与地球物理研究所浙江温州09 11 13 (杰青) 32 侯增谦1961 中国地质科学院河北藁城09 11 13(杰青)
城市道路等级分类
一、道路的组成及等级划分： （一）道路可分为：公路、城市道路、厂矿道路、林区道路及乡村道路。 1、公路的组成：线形组成、结构组成（路基、路面、桥涵、排水系统、隧道、防护工程、特殊构造物、交通服务设施） 2、城市道路组成 3、道路工程的主体是路线、路基（包括排水系统及防护工程等）和路面三大部分。 （二）道路等级的划分： 1、公路的等级划分：高速公路、一、二、三、四级5个等级。 2、城市道路的等级划分：快速路、主干路、次干路、支路。 二、路基 1、定义：路基是按照路线位置和一定技术要求修筑的作为路面基础的带状构造物。 2、分类： 填方路基（路堤）：填土路基、填石路基、砌石路基、护肩路基、护脚路基 挖方路基（路堑）：土质挖方路基、石质挖方路基 半挖半填路基 3、组成：路肩、边坡、排水设施、挡土墙。 4、对路基的要求： 1）有足够的强度：在车辆荷载、路面及路基自重作用下，变形不超过允许值。 2）有足够的整体稳定性：在行车及自然因素作用下，不发生滑坡、塌陷等。 3）有足够的水温稳定性：在水温变化时，路基强度变化小，不产生翻浆。 三、路面 （一）结构组成： 路层：是直接与行车及大气接触的表面层，应具有足够的结构强度、良好的温度 稳定性、耐磨、抗滑、平整和不透水。 基层：承受面层传来的垂直荷载并扩散至垫层和路基中。 垫层：在基层和土基之间的结构层，起隔水、防冻及扩散传力作用。 （二）度与路面排水 （三）路面的等级与分类 1、路面等级：高级路面、次高级路面、中级路面和低级路面。 2、路面类型： （1）路面基层的类型：按照现行规范，基层（包括底基层）可分为无机结合料稳定类和粒料类。 无机结合料稳定类有： 1)水泥稳定土基层：不应用作高级沥青路面的基层，在高速公路和一级公路的面板下 2)石灰稳定土基层：适用于各级公路路面的底基层，可作二级以下的公路的基层，但不应用作高级路面的基层。
化学-全国名校2020年高三6月大联考考后强化卷(新课标Ⅰ卷)(考试版)
化学试题 第1页（共8页） 化学试题 第2页（共8页） ………………○………………内………………○………………装………………○………………订………………○………………线………………○……………… ………………○………………外………………○………………装………………○………………订………………○………………线………………○……………… … 学校：______________姓名：_____________班级：_______________考号：______________________ 绝密★启用前 全国名校2020年高三6月大联考考后强化卷（新课标Ⅰ卷） 化 学 本卷满分100分，考试时间50分钟。 注意事项： 1．答卷前，考生务必将自己的姓名、准考证号填写在答题卡上。 2．回答选择题时，选出每小题答案后，用铅笔把答题卡上对应题目的答案标号涂黑。如需改动，用橡皮擦干净后，再选涂其他答案标号。回答非选择题时，将答案写在答题卡上。写在本试卷上无效。 3．考试结束后，将本试卷和答题卡一并交回。 可能用到的相对原子质量：H 1 C 12 N 14 O 16 Mg 24 Al 27 S 32 Cr 52 Fe 56 Ni 59 Ga 70 一、选择题：本题共7个小题，每小题6分，共42分。在每小题给出的四个选项中，只有一项是符合题目 要求的。 7．化学与生活密切相关，下列说法错误的是 A ．汉代烧制出“明如镜、声如磬”的瓷器，其主要原料为粘土 B ．雾霾天气会危害人类的健康，“雾”和“霾”的分散质微粒不同 C ．电热水器用镁棒防止内胆腐蚀，原理是牺牲阳极的阴极保护法 D ．煎炸食物的花生油和牛油都是可皂化的饱和高级脂肪酸甘油酯 8．化合物c 的制备原理如下： 下列说法正确的是 A ．该反应为加成反应 B ．化合物a 中所有原子一定共平面 C ．化合物c 的一氯代物种类为5种 D ．化合物b 、c 均能与NaOH 溶液反应 9．我国科研人员研究了在Cu?ZnO?ZrO 2催化剂上CO 2加氢制甲醇过程中水的作用机理，其主反应历程如 图所示(H 2→*H+*H)。下列说法错误的是 A ．向该反应体系中加入少量的水能增加甲醇的产率 B ．带*标记的物质是该反应历程中的中间产物 C ．二氧化碳加氢制甲醇的过程中原子利用率达100% D ．第③步的反应式为*H 3CO+ H 2O→CH 3OH+*HO 10．下列做法能达到相应实验目的的是 A B C D 验证浓硫酸具有脱水性和强氧化性 通过对照实验说明官能团对有机物的化学性质起决定作用 制取Cl 2 电泳实验证明Fe(OH)3胶体粒子带电 11．2019年5月，香港理工大学宣布研发出超柔软高效能织物[柔性S/ HPCNF (分级多孔碳纳米纤维)]锂空 气二次电池，在 S/ HPCNF 织物上均匀地沉积铜和镍，取代一般锂电池表面的金属箔，以提高柔软度，电池的工作原理如图，下列有关说法错误的是 A ．放电时，O 2在阳极区发生氧化反应 B ．沉积铜和镍是为了增强织物的导电性 C ．充电时，Li +移向金属锂一极 D ．充电时，阳极的电极反应为Li 2O 2?2e ? O 2↑＋2Li + 12．一种用于合成治疗免疫疾病药物的物质，其结构如图所示，其中X 、Y 、Z 、Q 、W 为1~20号元素且 原子序数依次增大，Z 与Q 同主族，Q 和W 的简单离子具有相同的电子层结构。下列叙述正确的是 A ．WX 是共价化合物 B ．原子半径：W>Z>Y C ．最高价氧化物对应的水化物的酸性：Q