Algorithms, Mehanisms and Proedures for the Computer-aided Projet Generation System Anton O. Butko 1*, Aleksandr P. Briukhovetskii 2, Dmitry E. Grigoriev 2# and Konstantin S. Kalashnikov 3 1 Department of Tehnologies of the Integrated Computer-aided Systems, Mosow Aviation Institute (National Researh University), Volokolamskoie highway, 4, Mosow 125993, Russia. 2 Department of Radio Engineering Devies and Antenna Systems, National Researh Institute Mosow Power Engineering Institute, Krasnokazarmennaya st., 14, Mosow 111250, Russia. 3 Department of Radio Physis, Voronezh State University, Universitetskaya sq., 1, Voronezh 394018, Russia. *Corresponding author 2# Orid: 0000-0002-0160-0724 Abstrat We suggest a simple, fast algorithm for the automated projet generation based on the data integrated in the invariant strutural-parametri model. The sample system is built (based) on this algorithm, and it inludes projet management software and interfae module. It allows user to prepare and ompose fully desribed projet within the Mirosoft Projet appliation. There is no manual data input or onversion needed. This solution an be used while dealing with any task requiring proess modeling and optimization, for example, data ompression, et. Keywords: Projet management, Automation, programming, Data onversion, Data ompression, Data transformation, Modeling, Strutural modeling, Parametri modeling. INTRODUCTION System for the automated generation of projets is a software produt designed for omputer-aided drafting (ompiling lists and tasks) for the purpose of alulation and optimization of any proess, inluding the prodution ones (organizational and tehnial solutions OTR). The system inludes several modules and is based on the invariant model provided by a struture-parametri modeling (SPM). To desribe the model, one uses linguisti apabilities and translators. The output is a omplete set of desriptions of the projet, whih an be loaded, for example, into Mirosoft Projet. This eliminates the need for manual data migration and settlement. A key element of the system is a software interfae for onverting data from the strutural-parametri model. The results of this development are intended to be applied for both omposition and desription of the varied data ompression proesses. THE MATHEMATICAL INTERPRETATION Let us onsider the mathematial interpretation of the funtional aspet of the system interfae. The formalization is arried out using the fundamental foundations of mathematis [1, 2] and set theory [3, 4]. The results of these studies lead us to the onlusion that the proess of the projet generation may be presented at a set-theoreti level, by the model ombining the three main interrelated mehanisms: where П(С)=П I (С)П II (С)П III (С), П I (С) produes analysis of the objet and extrats the design data on the basis of the invariant produt model; П II (С) onverts the objet struture on the basis of the tehnial solution; П III (С) performs both proessing and transformation of resoures (labor and material) and loads the obtained onversion results into the projet management system in order to omplete the projet. Let us present a desription of the mehanisms. П I (С)=A, N, B, R AN, G C, I. Here A is a set of model elements, N is a set of model parameters; B set of link sets (may be interpreted as R AA ); R AN are the relations whih determine the affiliation of parameters to elements; G C is a proess protool; I is a set of proedures that support proessing elements, parameters and relations in the model. We an see that the proedures to be determined will inlude software tools for the operation with the strutural-parametri database, solving the following ruial tasks: extrating elements, extrating parameters, proessing links. П II (С)=A, N, B, R AN, G C, II. (1) Here A is a set of model elements; N is a set of model parameters; B is a set of link sets; R AN are the relations whih determine the affiliation of parameters to elements; G C is a proess protool; II is a set of proedures that support the mehanisms of transformation of the olleted projet data into the import format of Mirosoft Projet. The omposition of the proedures defined in the II = ( f З, f С ), (2) 14199
where f З is the funtion of the task list reation, f С is the funtion of the formation of the system of relations. П III (С)=P, Q, L, R AN, G C, III. (3) Here P is a set of projet tasks; Q is a set of projet resoures; L is a set of task relationships sets (an be interpreted as R PP ); R PQ are the relations determining the affiliation of resoures to the tasks; G C is a proess protool; III is a set of proedures that support the mehanisms of the generated data loading into the projet management system: III = ( f Р, f 1,, f n ), where f Р is a funtion of the resoures proessing and loading. It is predetermined that these resoures will also inlude the proedures preparing and reording the import files f 1,,f n (the value of n depends on the funtionality of the interfae). Thus, as it has been previously mentioned, the struture of the interfae inludes three main mehanisms together solving the task of the projet generation. The most important among them is the mehanism of data onversion, so now we pass to the study of the basi funtions that it involves. FORMING THE TASK LIST The SPM apparatus analysis shows that the struturalparametri model in its relation to its onditions and tasks an be represented as this set of omponents: M SPM = (A, N, R AA, R AN ), (4) where A is a set of model elements, N is a set of model parameters, R AA are the relations (links) between the elements, R AN are the relations determining the affiliation of the parameters to the elements. We desribe the sets inluded in the SPM as A = { a 1,, a n }, (5) where a i is a single model element defined in the ELEMENTS blok (of the model); N = { n 1,, n n }, (6) where n i is a single parameter of the model defined in the PARAMETERS blok (of the model). The model of the projet resulting from the interfae operation an be desribed by the set of the following omponents: M P = (P, Q, R PP, R PQ ), Desription of the sets inluded into the projet then takes the form of where p i is a single projet task; P = { p 1,, p n }, (7) Q = { q 1,, q n }, where q i is a single unit of the resoures involved in the projet, as defined in the "LIST OF RESOURCES". However, given the arrangement that the SPM elements system and the tasks in the projet have, it is more onvenient to express these omponents through the set where eah element itself is a set of elements of the model, or of the projet, belonging to the link unit: M SPM = (A, N, B, R AN ), M P = (P, Q, L, R PQ ). All the model elements, as well as its parameters, possess ertain properties. In the present ase, the set of properties an be generally presented as a i A (F(a i) = (F 1 F n)), where F i is a property of an element. In order to perform orret work of mehanism П I, the item onverted into a projet task neessarily differs from the rest of the model elements. Firstly, it has to be a part of the solution, and, seondly, it is to be marked with a parameter marker. These are the properties needed. Figures 1 and 2 show how the elements of the model are seleted for their subsequent inlusion in the list of tasks. Let now desribe the set of properties with the help of the table from Figure 1. This is example for a model with 5 elements and 7 possible properties (the dot indiates whether the element has a property). So, there we an find the possible model elements properties, the two of them marked as the ones whose presene makes the ourrene of the element in the future tasks list possible. The tables a and b in Figure 2 present the approximate set of elements and their properties. The table a presents all the elements, and the table b implements the elements seletion. f 1 f 2 f 3 f 4 f 5 f 6 f 7 F(A) Figure 1: Charateristis of the task-element where P is a set of projet tasks, Q is a set of resoures, R PP are the relations (links) between the tasks, and R PQ are the relations determining the belonging of the resoures to the tasks. 14200
f 1 f 2 f 3 f 4 f 5 f 6 f 7 a 1 a 2 a 3 a 4 a 5 a) f 1 f 2 f 3 f 4 f 5 f 6 f 7 a 2 a 3 b) Figure 2: The tables of the harateristis and of the results of the elements seletion: a) harateristis of the SPM elements; b) the result of the elements seletion From here we selet the ondition for the ourrene of a model element in a future projet as a task: F(A) = (F 3, F 4 ). The solution produed by means of the speified tables is A = (a 2, a 3). That eah element possesses some speifi properties an be desribed by a Boolean matrix of the following form: ij A F A F1 11 12 n1 F2 12 22 n2 Fm 1m a1 2m a2 nm an The parameter takes the value 0 or 1 depending on the properties of the element: ij 1, if Fi F( ai ), 0, if Fi F( ai ). As the seleted properties may belong to the multiple elements, the olletions of items for eah property should be represented as a separate set K = { k 1,, k n }, where we an see that k i has a parameter marker. Then k i : k i A, but it may be that а i K, A: A K, K: F(A) F(K). And the set of the elements inluded in the solution is E = { e 1,, e n }, where e i is the item, for whih the following ondition is fulfilled: e i ELRESH (the list of solution items). Then e i : e i A, but it may be that а i E, A: A E, E: F(A) F(E). As we see, F(A) ontains all the model elements, inluding support and geometri elements, their transmission into the projet management system not required. Thus, F(E) ontains only those elements that need to be transferred to the projet for planning. Finally, we selet a set of elements possessing both properties: D = { d 1,, d n }, where d i is the element with the parameter marker and is inluded in the solution. Then d i : d i A, but it may be that а i D, A: A D, D: F(A) F(D). There F(A) ontains all the model parameters, inluding support and geometri parameters, their transmission into the projet management system not required. Thus, F(D) ontains only those parameters that need to be transferred to the projet for planning. Hene, the set D, as well as K and E, is a subset of A and is formed by the intersetion of the sets K and E: D = K E, F(D)=F(K) F(E). Now we define the ondition for the seletion of the elements to be inluded into the projet: a i D ((a i E)( a i K)) or a i D (F(a i) = F 3(a i) F 4(a i)). As a result, the task of produing the projet task list for the interfae is redued to the desription of the funtion f З: p i = f З(d i). This feature is to be inluded into the mehanism of the SPM data onversion into the Mirosoft Projet data. CREATING A LIST OF RESOURCES The next task is a transfer of resoures. It involves the deision similar to the presented above. As resoures in the SPM are represented by the parameters of a ertain type, we implement additional set V: V = { v 1,, v n }, where v i is the parameter of the "resoure" type. Then 14201
v i : v i N, but it may be that v i N, N: N V, V:F(N) F(V). We define all the types of the parameters via the harateristis table (Fig.3). f 1 f 2 f 3 f 4 f 5 f 6 f 7 F(N) Figure 3: Charateristis of the task-element The properties we are interested in are the parameter types "working resoures" and "material resoures". Thus, we an determine the ondition for sampling the parameters to be inluded into the projet as a ertain type of resoures in the following way: n i V (F(n i) = F 6(n i) F 7(n i)). As a result, the task of forming projet resoures for the interfae is redued to the desription of the funtion f Р: q i = f Р(v i). This feature will be inluded in the onversion mehanism of SPM data into the projet data for Mirosoft Projet. THE MODEL AND ALGORITHM In general, we an onlude that the task of developing an interfae an be redued to the implementation of the mehanisms presented above and to the appliation of the modern software development tools. Thus, our three mehanisms will work in the following way: П I (С) the mehanism is based on the built-in proedures for the SPM operating, allowing aessing the elements and parameters of the model in order to retrieve the neessary data; П II (С) the mehanism is based on the algorithms for the data transformation implemented in the environment of Mirosoft Visual Studio C++, its basi sheme onsists in an iterative proessing of the arrays of elements extrated from the model; П III (С) this mehanism works on the basis of the algorithms for the data transformation, implemented in the environment of Mirosoft Visual Studio C++. The basi sheme onsists in an iterative proessing of arrays of parameters (resoures), extrated from the model. The transfer of resoures into the projet management system provides a maro reated by means of VBA. Therefore, the system interfae will inlude three key algorithms: 1. Algorithm for data analysis and extration; the algorithm represents a yle with sequential seletion of elements and parameters and their further analysis (verifiation onditions), aiming at the generation of a projet struture. 2. Algorithm for data onversion and transfer; the algorithm represents a yle of the serial onversion of the obtained struture of the projet into a format suitable for its input into the projet management system. 3. Algorithm for the generation of the list of resoures; the algorithm is a ylial proessing of resoures parameters for resoures alloation (by marker validation) and the generation of their list in the format of the projet management system. Before a software system development, the study is onduted, in order to identify the possible problems and peuliarities, taking into aount the available software produts that an be used as the omponents of the system. Depending on the a priori aepted onditions, there an be several design algorithms. We should onsider the most important fators that, in fat, are the deisive riteria when hoosing a suitable variant for the algorithm implementation. First suh fator is the presene or absene of tools to work with formats and internal funtions of the automated system (AS) and detailed guidelines for their use. In order to develop an algorithm, whih works with the formats of different systems, first of all, you must either learly represent the struture and the form in whih tehnial information is stored in formats of AS, or have a fully desribed and extensively worked on software tools that works with this format. Quite often suh tools and guidane on their appliation is provided by the developer together with the AS, usually the library funtions are ompiled into the standard Windows format DLL Dynamially Linked Library. In order to work with these omponents programmer simply onnet them to the projet and immediately gets the opportunity to work with different internal funtions of AS [5]. In this ase, the task of implementing the interfae beomes simpler, and the work is redued mainly to the maintaining software data onversion. Different situation arises, if the AS produer does not provide the doumentation and other means neessary for external developers to operate the features and formats of AS and just gives a detailed desription of all the formats and aess methods. The system may support only its own speifi and losed formats for information import/export and not the ommon ones. In this ase, the programmer would have to write all the funtions performing the needed information/data extration and proessing, the solution that may require the appliation of heuristi methods. This will inevitably affet the ost of the designed produt, in some ases serving as a possible reason for the rejetion of the hoie of this partiular AS, with subsequent revision of the desired system integration method. 14202
The seond fator is whether the AS supports one or more of ommon import formats, whose detailed ontent desription and the struture, as well as their methods and tools, are available for developers. As a rule, for them there are already many third-party software proessing tools, inluding the open soure ones. The next option is: the AS produer does not provide any supporting tools and data for external developers, but the system supports additional import formats, ontent and desription of whih are widely known and available for the programmer, as a DXF format for CAD systems, for example. In this situation, the most diffiult work is redued to the transformation of the data/information and providing its auray and integrity [6]. If the support of ommon formats is missing and there is no software apable of operating the internal funtions of the system, i.e. it is losed as its format, one would have to make a diffiult deision, or to selet a different AC, or a different method of integration of suh systems. The third fator onsists in the size of the budget and the time frame alloated for the projet. They radially affet the hoie of methods and means of integration, espeially for Russian ompanies. Now, on the basis of the above stated, we an present a general desription of the basi options for the ontent of the algorithm onstruting the software interfae, aiming at the integration of the seleted systems. These options differ from one another by presene or absene of ertain operations and by the sequene of operations in the algorithms they provide. Option 1. Diret onversion of the formats from one to another is arried out without running the AS. The interfae works with the file of the system from whih you import by means of the internal funtions of this very system. The result is stored in the importing system s format. In this ase, no AS installed on a omputer is required, whih is one of the advantages of this option. Option 2. The formats are onverted from one to another through the intermediate format, without running the AS. This option presupposes that a ertain external auxiliary onverter is used, performing the transformation of data from the intermediate format into the importing system s format. In this ase, there is also no need to install AS. Option 3. Data onversion triggers one of the AS and utilizes its internal funtions. The interfae operation is only possible on the omputer with the ertain AS and the partiular API funtions library installed (if the latter is not supplied with the interfae already). Option 4. The data onversion triggers both systems and utilizes their internal funtions. As it has been previously mentioned, as a system data soure the SPM Modeler is seleted, and its appliability is determined by the following reasons: SPM is invariant to the semantis of the data in the ontext of the speifi objet modeling; the Modeler allows generating geometri models of omplex topology; SPM an be used in "assembly diagrams" nodes models; SPM S(A) is based on the strutural properties of the systems and an be used in data formulation for the elements of the simulation objet A, the properties of F and the relations between the elements and the properties of R; matrix apparatus and system of relations are the algorithmi basis for the solution of a wide range of tasks by applying a modular priniple of information management during the models proessing; SPM provides extensive features for parameterization and struturing. To illustrate the operation of the system, Mirosoft Projet is seleted as the projet management system, as it is haraterized by a universal availability and a great variety of funtions implementing the tasks regulated in the MRP-I and MRP-II. The hoie in favor of this lass of systems is the result of the suggestion aepted to introdue suh AS as the omponent of the integrated solution, instead of the ERP, that, for the domesti enterprises, appears to be hard to implement and apply. The more so, as the struture of the information oming as the original data in SPM is almost equivalent to the one in ERP, or PDM, systems. Theoretially, the hoie of the AS is not a key fator, signifiantly influening the appliability of the suggested sheme (PS (Modeler) interfae PM (PDM, ERP in existing solutions)), due to the invariane of the latter relative to the speifi systems. Thus, the solution (the proedure) is expeted to have suffiient adaptability, at least within our field of study and researh. The results of this analysis based on the projet operation tasks and the AS seleted, a deision is made to develop the algorithm using the ombined approah involving the following interfae operation sequene: the soure data is extrated by means of the SPM speial internal funtions, used for the external aess to the SPM format without triggering the entire omplex; and then the information is strutured, verified and onverted to the import format of the projet management systems. Part of the data is generated in intermediate format, for its subsequent loading into the system by means of a speial maro integrated in the projet management system [7]. The resulting import file is opened in Mirosoft Projet environment and the generation of the projet framework begins, followed by the loading of resoures and the 14203
generation of the resoure table, until the projet generation is ompleted. The results of the projet design are presented as a high-level algorithm of the interfae module operation, desribed following the IDEF0 methodology [8, 9]. The orresponding hart is shown in Fig. 4, 5. Fig. 4b presents the deployment of the root diagram showing the basi stages of the projet generation, implemented by means of the three interfae mehanisms desribed above. Fig. 5a, presents the deployment of the first blok of the first level diagram, demonstrating the main stages of the extration proess. Fig. 5b presents the deployment of the seond blok of the first level diagram, showing the basi steps of the data onversion proess. Fig. 5 shows the deployment of the third blok of the first level diagram, illustrating the main stages the projet data loading into the Mirosoft Projet takes. a) b) Figure 4: Upper level diagrams: a) root diagram; b) first level diagram (shows usage of П I П II П III mehanisms) a) 14204
b) ) Figure 5: Bloks deployment for the first level diagram: a) diagram 1, seond level; b) diagram 2, seond level (shows usage of f 3, f P, f C funtions); ) diagram 3, seond level As we an see, the diagrams below present the harts that depit the most important stages of the proess of the interfae operation and, in fat, onstitute together the oneptual sheme of the projet generation proedure. П II (С) operates within bloks 7, 10, 11; П III (С) operates within blok 13. Fig. 6 shows high-level algorithm for the proess of data transition from SPM to Mirosoft Projet: П I (С) operates within blok 6; 14205
Figure 6: High-level algorithm sheme The diagrams present the main stages and bloks of the proess of the interfae operation. Most of these bloks inlude yli algorithms for the operation with the arrays of elements and parameters of the model to be onverted into tasks and projet resoures. The suggested solution was tested as a part of the orporate information system for management and planning, deployed at a spae plant. The system is based on the omplex modeling produt (SPM), with the Mirosoft Projet environment managing the projet applied. This omplex was used for the tehnial produts manufaturing planning. A trial period shows signifiant redution of the data input time and errors quantity. The period needed for the data entry into the projet management system is redued almost by two times. But there are still some restritions enountered during the testing: some versions of the Mirosoft Projet appliations annot support large models that ontain huge numbers of elements (from 100 000 to 1 000 000). Given this, we have to test the models of the middle and small sizes. The suggested solution allows avoiding manual entry of data into the projet management system, and thus the number of errors is redued. We expet a signifiant redution in the time it takes to build the projets for planning through the automated data transfer from the modeling system. CONCLUSION The suggested simple and fast digital algorithm onverting data for projet generation allows reating and managing various proess models, in order to find optimal (balaned) solution. Projets an be loaded into any Projet Management system for the subsequent analysis and optimization. Suggested solution provides the possibilities to avoid multiple errors produed by the manual data input and onversion. 14206
ACKNOWLEDGEMENT This work was supported by the Ministry of Eduation and Siene of the Russian Federation (projet No. 14.577.21.0163). REFERENCES [1] Gorbatov, V.F., 2000, Fundamentals of Disrete Mathematis (in Russian), Nauka, Fizmatlit, Mosow. [2] Chenin, P., Cosnard, M., Gardan, Y., et al., 1985, Mathematiques et САО: In 2 Volumes, Hermes Publishing, New Castle. [3] Mostowski, A., 1969, Construtive Sets with Appliations, North Holland PWN, Amsterdam Warszawa. [4] Jeh, T.J., 2003, Set Theory, Springer-Verlag, Berlin. [5] Blagodatskih, V.A., Engibaryan, M.A., Kovalevskaya E.V., et al., 1995, Eonomis, Development and Use of PC Software (in Russian), Mosow, Finansy i Statistika. [6] Hotyashov, E.N., 1995, Fundamentals of Automati Data Proessing Systems Engineering (in Russian), Finansy i Statistika, Mosow. [7] Martin, P., and Tate, K., 2001, Getting Started in Projet Management, Wiley, New Jersey. [8] Integrated Information Support System. Information Modeling Manual IDEF0-Extended. ICAM Projet Priority 6201, General Eletri Company, New York, 1985. [9] Integrated Computer-Aided Manufaturing (ICAM) Arhiteture. Part II. Volume IV Funtion Modeling Manual (IDEF0), SoftTeh. In., Waltham, 1981. 14207