Power grid project planning pattern recognition method and system based on time series feature matching

Abstract

A method and system for recognizing patterns in power grid project planning based on time series feature matching. The method first constructs a historical project progress curve pattern library, extracts and weights the project progress time series features; then constructs a membership function model based on the weights, matches the project progress features, and forms an optimal matching set; then defines and solves a dynamic time warping path model, optimizes the path through a distance matrix, and generates a new project progress curve sequence; finally, the sequence and the best matching set features are integrated to obtain a final project progress curve pattern; the invention significantly improves the prediction accuracy by accurately capturing and weighting the project progress time series features, and optimizes the recognition matching algorithm through the dynamic time warping technology, thereby enhancing the rationality and accuracy of the comparison between project data, and at the same time reduces the interference of human factors on the recognition results, thereby ensuring the objectivity and accuracy of the recognition results.

Description

Technical Field

[0001] The invention relates to a project planning pattern recognition method, belongs to the field of power grid infrastructure progress recognition, and in particular to a power grid project planning pattern recognition method and system based on time series feature matching. Background Art

[0002] There are a large number of different project progress curve patterns for massive power grid infrastructure projects due to their different characteristics such as project nature, construction scale, voltage level, etc. The progress curve of power grid infrastructure projects is a cluster of complex time series with nonlinear and non-stationary characteristics. It reflects the construction progress of an infrastructure project and its dynamic changes with the advancement of the construction period. It has important guiding significance in the process of identifying planning patterns for massive power grid infrastructure projects.

[0003] The progress curve of power grid infrastructure projects affects the optimization of power grid infrastructure planning projects. For the optimal combination of the best projects, whether all power grid projects can be completed on schedule within the planned construction period depends largely on the accuracy of the recognition of the project progress curve pattern. However, the traditional recognition of the progress curve pattern of power grid infrastructure projects often relies on the subjective experience of experts, fails to give full play to the intrinsic value of the massive historical infrastructure project progress data in the power grid information system, ignores the time series characteristics of the progress curve of power grid infrastructure projects, and is difficult to provide scientific and objective quantitative decision support for the subsequent optimization and dynamic time series deduction of power grid infrastructure planning projects. Therefore, there is an urgent need for an objective and accurate means to solve the above-mentioned defects in the existing technology. Summary of the invention

[0004] The purpose of the present invention is to overcome the above-mentioned defects and problems existing in the prior art and to provide an objective and accurate method and system for identifying power grid project planning patterns based on time series feature matching.

[0005] To achieve the above objectives, the technical solution of the present invention is: a method for recognizing a pattern of power grid project planning based on time series feature matching, comprising:

[0006] S1. Build a historical project progress curve pattern library, obtain project progress timing characteristics, and assign weights to each project progress timing characteristics according to the degree of discreteness;

[0007] S2. Based on the weights of the project progress timing characteristics, a project progress curve membership function model is constructed, and the project to be identified is matched with the project progress timing characteristics in the project progress curve membership function model to obtain the best matching set of project progress timing characteristics;

[0008] S3. Based on the best matching set of project progress timing characteristics, define a distance matrix of the project progress curve, and based on the distance matrix, define a first dynamic time warping path and conditions to obtain a dynamic time warping path model;

[0009] S4. Solve the dynamic time warping path model to obtain the second dynamic time warping path, and expand the project progress curve according to the second dynamic time warping path to obtain a new project progress curve sequence, and calculate to obtain a comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, merge all project progress curves with all project progress timing features in the best matching set of project progress timing features to obtain the final project progress curve pattern.

[0010] The step S1 specifically includes:

[0011] S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library;

[0012] S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics;

[0013] S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows:

[0014] ;

[0015] in: Indicates The weights of the project schedule timing characteristics, Represents the first The sample variance of the project schedule timing characteristics.

[0016] The step S2 specifically includes:

[0017] S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance;

[0018] S22, inputting the project to be identified into the project progress curve membership function model to match it with the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics.

[0019] The step S22 specifically includes:

[0020] S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified;

[0021] S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them;

[0022] S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model;

[0023] The expression of the project progress curve membership function model is as follows:

[0024] ;

[0025] in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project;

[0026] S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223;

[0027] S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

[0028] The step S3 specifically includes:

[0029] S31. Assume that the project progress curve with the highest matching degree in the best matching set of project progress timing characteristics is , The sequence length is ; Assume that the progress curve of any other project in the best matching set of project progress timing characteristics is , The sequence length is ; Define the project progress curve and The distance matrix is ​​as follows:

[0030] ;

[0031] in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves;

[0032] S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows:

[0033] ;

[0034] in: represents the starting point of the first dynamic time warping path; represents the end point of the first dynamic time warping path; represents the length of the first dynamic time warping path, which satisfies , ; represents the first dynamic time warping path No. Project progress curve at each location sequence points and sequence points Match;

[0035] S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows:

[0036] ;

[0037] in: represents the first dynamic time warping path Middle Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the first dynamic time warping path Two adjacent points in the .

[0038] The step S4 specifically includes:

[0039] S41. Solve the first dynamic time warping path model based on the constrained dynamic time warping algorithm to obtain a project progress curve and The corresponding second dynamic time warping path ;

[0040] S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ;

[0041] S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is equal in length to the second dynamic time warping path;

[0042] The expression of the comprehensive project progress curve sequence is as follows:

[0043] ;

[0044] in: and Represent the projects to be identified and the project progress curve respectively and The degree of membership of the progress curve of the corresponding project; Represents the best matching set of the project to be identified and the project schedule characteristics Sum the membership of the progress curves of all projects in;

[0045] S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; The expression is as follows:

[0046] ;

[0047] in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point positions A collection of; express The number of elements in the collection; Indicates the first term curve;

[0048] S45, determining whether the maximum number of iterations is met;

[0049] If satisfied, the final project progress curve model is output;

[0050] If not satisfied, the project progress curve Updated to , and repeat steps S41-S45 until the The best matching set of project progress timing characteristics corresponding to the projects to be identified All the project progress curves in the project are integrated to obtain the final project progress curve model.

[0051] A power grid project planning pattern recognition system based on time series feature matching, the system comprising:

[0052] The project progress timing feature acquisition module is used to build a historical project progress curve pattern library, obtain the project progress timing features, and assign the weight of each project progress timing feature according to the discrete degree;

[0053] A project progress timing feature best matching set building module is used to build a project progress curve membership function model based on the weight of the project progress timing feature, and match the project to be identified with the project progress timing feature in the project progress curve membership function model to obtain the best matching set of project progress timing features;

[0054] A dynamic time warping path model building module is used to define a distance matrix of a project schedule curve based on the best matching set of project schedule timing characteristics, and to define a first dynamic time warping path and conditions based on the distance matrix to obtain a dynamic time warping path model;

[0055] The project progress curve pattern acquisition module is used to solve the dynamic time warping path model, obtain the second dynamic time warping path, and expand the project progress curve according to the second dynamic time warping path to obtain a new project progress curve sequence, and calculate the comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, all project progress curves are merged with all project progress timing features in the best matching set of project progress timing features to obtain the final project progress curve pattern.

[0056] The project progress timing feature acquisition module acquires the project progress timing features and their weights according to the following steps:

[0057] S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library;

[0058] S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics;

[0059] S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows:

[0060] ;

[0061] in: Indicates The weights of the project schedule timing characteristics, Represents the first The sample variance of the project schedule timing characteristics.

[0062] The project progress timing feature best matching set construction module obtains the project progress timing feature best matching set according to the following steps:

[0063] S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance;

[0064] S22, inputting the project to be identified into the project progress curve membership function model to match the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics;

[0065] The step S22 specifically includes:

[0066] S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified;

[0067] S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them;

[0068] S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model;

[0069] The expression of the project progress curve membership function model is as follows:

[0070] ;

[0071] in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project;

[0072] S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223;

[0073] S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

[0074] The dynamic time warping path model construction module constructs a dynamic time warping path model according to the following steps:

[0075] S31. Assume that the project progress curve with the highest matching degree in the best matching set of project progress timing characteristics is , The sequence length is ; Assume that the progress curve of any other project in the best matching set of project progress timing characteristics is , The sequence length is ; Define the project progress curve and The distance matrix is ​​as follows:

[0076] ;

[0077] in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves;

[0078] S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows:

[0079] ;

[0080] in: represents the starting point of the first dynamic time warping path; represents the end point of the first dynamic time warping path; represents the length of the first dynamic time warping path, which satisfies , ; represents the first dynamic time warping path No. Project progress curve at each location sequence points and sequence points Match;

[0081] S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows:

[0082] ;

[0083] in: represents the first dynamic time warping path Middle Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the first dynamic time warping path The conditions for two adjacent points to be satisfied;

[0084] The project progress curve mode acquisition module acquires the final project progress curve mode according to the following steps:

[0085] S41. Solve the first dynamic time warping path model based on the constrained dynamic time warping algorithm to obtain a project progress curve and The corresponding second dynamic time warping path ;

[0086] S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ;

[0087] S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is equal in length to the second dynamic time warping path;

[0088] The expression of the comprehensive project progress curve sequence is as follows:

[0089] ;

[0090] in: and Represent the projects to be identified and the project progress curve respectively and The degree of membership of the progress curve of the corresponding project; Represents the best matching set of the project to be identified and the project schedule characteristics Sum the membership of the progress curves of all projects in;

[0091] S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; The expression is as follows:

[0092] ;

[0093] in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point positions A collection of; express The number of elements in the collection; Indicates the first term curve;

[0094] S45, determining whether the maximum number of iterations is met;

[0095] If satisfied, the final project progress curve model is output;

[0096] If not satisfied, the project progress curve Updated to , and repeat steps S41-S45 until the The best matching set of project progress timing characteristics corresponding to the projects to be identified All the project progress curves in the project are integrated to obtain the final project progress curve model.

[0097] Compared with the prior art, the present invention has the following beneficial effects:

[0098] The present invention discloses a method and system for recognizing a pattern of power grid project planning based on time series feature matching. The method first constructs a historical project progress curve pattern library, extracts and weights the project progress time series features; then constructs a membership function model based on the weights, matches the project progress features, and forms an optimal matching set; then defines and solves a dynamic time warping path model, optimizes the path through a distance matrix, and generates a new project progress curve sequence; finally, the sequence and the optimal matching set features are integrated to obtain a final project progress curve pattern; in application, the present invention significantly improves the prediction accuracy by accurately capturing and weighting the project progress time series features, and optimizes the recognition matching algorithm through the dynamic time warping technology, thereby enhancing the rationality and accuracy of the comparison between project data, and at the same time reduces the interference of human factors on the recognition results, thereby ensuring the objectivity and accuracy of the recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 It is a flow chart of the method steps of the present invention.

[0100] Figure 2 It is a flowchart of the method steps of the project progress curve matching strategy based on weighted Hamming distance membership in Example 1 of the present invention.

[0101] Figure 3 It is a schematic diagram of the dynamic time warping path of the conventional DTW algorithm in Example 1 of the present invention.

[0102] Figure 4 This is a schematic diagram of the dynamic time warping path of the constrained dynamic timing warping algorithm in Example 1 of the present invention.

[0103] Figure 5 It is a flow chart of the method steps for determining the final project progress curve model in the present invention.

[0104] Figure 6 It is a schematic diagram of a progress curve model for a 110kV urban network power transmission and transformation project for solving equipment heavy overload in Example 1 of the present invention.

[0105] Figure 7 It is a schematic diagram of a progress curve model of a certain 110kV urban grid in Example 1 of the present invention to meet the needs of a new load transmission and transformation project.

[0106] Figure 8 It is a schematic diagram of a progress curve model of a 110kV rural power grid power transmission and transformation project for solving equipment heavy overload in Example 1 of the present invention.

[0107] Fig. 9 It is a schematic diagram of a progress curve model of a 110kV rural power grid in Example 1 of the present invention to meet the needs of a new load transmission and transformation project.

[0108] Fig.10It is a schematic diagram of a progress curve model of a 110kV rural power grid power transmission and transformation project for eliminating safety hazards of obsolete equipment in Example 1 of the present invention.

[0109] Fig.11 It is a schematic diagram of the system structure of the present invention.

[0110] Fig.12 It is a schematic diagram of the device structure of the present invention.

[0111] In the figure: project progress timing feature acquisition module 1, project progress timing feature best matching set construction module 2, dynamic time warping path model construction module 3, project progress curve pattern acquisition module 4, processor 5, memory 6, computer program code 61. DETAILED DESCRIPTION

[0112] The present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0113] Embodiment 1:

[0114] In this embodiment, a company's information system stores a large number of historical project progress curves, which can be regarded as a historical project progress curve pattern library of power grid infrastructure projects, where each historical project progress curve is regarded as a historical project progress curve pattern. For a new power grid infrastructure project, after obtaining its corresponding project characteristics and basic characteristics, the time series characteristics of the power grid infrastructure project progress can be matched based on the distance metric. Each new infrastructure project sample matches one or several historical infrastructure projects in the power grid infrastructure project historical progress curve pattern library. This decision-making and prediction process is the fuzzy matching of "point to set".

[0115] The essential feature of the matching of the time series characteristics of the progress of power grid infrastructure projects is that the historical project progress curve patterns of several power grid infrastructure projects are known in advance. From the perspective of fuzzy set theory, the process of judging which project progress curve pattern each new power grid infrastructure project to be matched belongs to is fuzzy. There may be several infrastructure project progress patterns with high matching degree at the same time. This fuzzy identification of the progress curve of power grid infrastructure projects follows the maximum membership principle. That is, the historical project progress curve pattern library There are Historical project progress curve model , is a power grid infrastructure project to be identified; if the power grid infrastructure project to be identified Corresponding project progress curve membership , then the power grid infrastructure project needs to be identified Priority belongs to the historical project progress curve model , which can be considered as infrastructure projects to be identified Similar to historical infrastructure projects , using historical infrastructure projects Can represent infrastructure projects to be identified Based on the above maximum membership principle, a membership function model of the progress curve of power grid infrastructure projects can be constructed as a criterion for matching the infrastructure projects to be identified with historical infrastructure projects.

[0116] Based on the fuzzy matching principle described above, the specific steps are as follows:

[0117] See also Figure 1 , a power grid project planning pattern recognition method based on time series feature matching, comprising:

[0118] S1. Build a historical project progress curve pattern library, obtain project progress timing characteristics, and assign weights to each project progress timing characteristics according to the degree of discreteness;

[0119] The step S1 specifically includes:

[0120] S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library;

[0121] S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics;

[0122] In this embodiment, it is necessary to select the project features required for matching the progress timing characteristics of the power grid infrastructure project. In the present invention, these feature sets are defined as the power grid infrastructure project progress timing feature vector ,The project progress timing characteristics include the project commissioning time and the reasonable construction period. Based on the historical project progress curve pattern library in the power grid information system, the project progress timing characteristics screened and determined include the total investment in the project feasibility study, voltage level, project nature, engineering attributes, construction scale and regional attributes.

[0123] Among them, voltage levels include 35kV, 110kV, 220kV, and 500kV; the nature of the projects includes substation engineering projects, line engineering projects, and transmission and transformation engineering projects; the project attributes include solving heavy overloads, meeting new loads, strengthening grid structures, supporting electric railways, supporting substation transmission, and eliminating safety hazards of old equipment; the construction scale includes project substation capacity and line length; the regional attributes include urban grid projects and rural grid projects.

[0124] Since the dimensions and magnitudes of the project schedule time series characteristics in the historical project schedule curve pattern library are different, if the value of a project schedule time series characteristic is very large, it may cause the project schedule time series characteristic matching result to ignore the influence of the distance between other project schedule time series characteristics on the membership of the total project schedule curve, making the project schedule curve membership unable to measure the most matching historical project schedule curve pattern well. Therefore, before establishing the project schedule curve membership function model considering the time series characteristics, it is necessary to normalize the project schedule time series characteristics, and the normalized data is between [0,1].

[0125] S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows:

[0126] ;

[0127] in: Indicates The weights of the project schedule timing characteristics, Represents the historical project progress curve pattern library The sample variance of the project schedule timing characteristics.

[0128] In this embodiment, the greater the degree of discreteness, the more obvious the fluctuation range of the project progress time series characteristics, and the more sensitive the matching process of the power grid infrastructure project progress time series characteristics. This shows that the historical project progress curve pattern library itself is dynamically expandable and can adapt to the continuous entry of new power grid infrastructure projects. As the number of power grid infrastructure projects in the library increases, the accuracy of the matching of the power grid infrastructure project progress time series characteristics will continue to improve.

[0129] S2. Based on the weights of the project progress timing characteristics, a project progress curve membership function model is constructed, and the project to be identified is matched with the project progress timing characteristics in the project progress curve membership function model to obtain the best matching set of project progress timing characteristics;

[0130] Furthermore, the step S2 specifically includes:

[0131] S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance;

[0132] S22, inputting the project to be identified into the project progress curve membership function model to match it with the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics.

[0133] The step S22 specifically includes:

[0134] S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified;

[0135] S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them;

[0136] S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model;

[0137] In this embodiment, the matching of the time series characteristics of the progress of the power grid infrastructure project is essentially the comparison of the membership of the project progress curve; the larger the value, the higher the degree of membership of the project to be identified to the historical progress curve pattern of a certain power grid infrastructure project. By analyzing the massive data of the normalized historical project progress curve pattern library, it is found that the distribution of the time series characteristics of the progress of the selected power grid infrastructure project is suitable for using the weighted Hamming distance to construct the project progress curve membership function model.

[0138] The expression of the project progress curve membership function model is as follows:

[0139] ;

[0140] in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project;

[0141] S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223;

[0142] S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

[0143] Thus, the degree of affiliation between the project to be identified and all historical projects in the historical project progress curve pattern library can be reliably quantified, and then the best matching set of project progress timing characteristics that highly matches the progress pattern of the project to be identified can be found. , as input to the recognition method.

[0144] In this embodiment, the method for obtaining the best matching set of project progress time series features through the above strategy is called a project progress curve matching strategy based on weighted Hamming distance membership. The specific steps are as follows: Figure 2 As shown. The matching of the time series characteristics of the progress of power grid infrastructure projects based on the weighted Hamming distance is a fuzzy recognition method. This method believes that the progress curves of power grid infrastructure projects with similar project progress time series characteristics can be matched through historical project samples; based on the weighted Hamming distance, a project progress curve membership function model considering time series characteristics is introduced to find one or several historical project samples in the historical project progress curve pattern library that are most similar to the current project progress time series characteristics, thereby achieving accurate matching of massive power grid infrastructure projects.

[0145] Based on the best matching set of project progress timing characteristics obtained in steps S1 and S2 Based on the curve fusion theory, the present invention constructs a dynamic time warping path model for matching multi-project progress curves, and further solves it through a restricted dynamic time warping algorithm to fuse the historical project progress curve patterns that are highly matched with the project to be identified, thereby forming a method for identifying massive project progress curve patterns. The details are as follows:

[0146] S3. Based on the best matching set of project progress timing characteristics, define a distance matrix of the project progress curve, and based on the distance matrix, define a first dynamic time warping path and conditions to obtain a dynamic time warping path model;

[0147] Furthermore, the step S3 specifically includes:

[0148] S31. Assume that the project progress curve with the highest matching degree in the best matching set of project progress timing characteristics is , The sequence length is ; Assume that the progress curve of any other project in the best matching set of project progress timing characteristics is , The sequence length is ; Define the project progress curve and The distance matrix is ​​as follows:

[0149] ;

[0150] in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves;

[0151] S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows:

[0152] ;

[0153] in: represents the starting point of the first dynamic time warping path; represents the end point of the first dynamic time warping path; represents the length of the first dynamic time warping path, which satisfies , ; represents the first dynamic time warping path No. Project progress curve at each location sequence points and sequence points Match;

[0154] S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows:

[0155] ;

[0156] in: represents the first dynamic time warping path Middle Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the first dynamic time warping path Two adjacent points in the .

[0157] In this embodiment, the first dynamic time warping path Satisfy the following properties:

[0158] 1) Boundary conditions: and , to ensure the project construction progress curve and All sequence points show the first dynamic time warping path matching the multi-project progress curves middle.

[0159] 2) Continuity: The first dynamic time-warping path for matching multiple project schedules Two adjacent points and satisfy and , indicating the project construction progress curve and The sequence points in can only be matched with adjacent points, and the first dynamic time warping path can be further obtained. The monotonicity of and ;in: .

[0160] S4. Solve the dynamic time warping path model to obtain the second dynamic time warping path, and expand the project progress curve according to the second dynamic time warping path to obtain a new project progress curve sequence, and calculate to obtain a comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, merge all project progress curves with all project progress timing features in the best matching set of project progress timing features to obtain the final project progress curve pattern.

[0161] Furthermore, the step S4 specifically includes:

[0162] S41. Solve the first dynamic time warping path model based on the constrained dynamic time warping algorithm to obtain a project progress curve and The corresponding second dynamic time warping path ;

[0163] In this embodiment, the first dynamic time warping path model can be solved by using a dynamic time warping algorithm (DTW). However, since the conventional DTW algorithm lacks restrictions on project construction progress curves and matching points, pathological dynamic time warping paths with matching of multiple project progress curves may occur, such as Figure 3 As shown. Figure 3 It can be seen that although the cumulative distance of the pathological dynamic time warping path matched by the multi-project progress curve is small, it deviates seriously from the diagonal line, and there is a project construction progress curve and The over-matching problem affects the generalization ability of the massive project progress curve fusion algorithm.

[0164] Therefore, the present invention adopts a restricted dynamic time warping algorithm to solve the dynamic time warping path model for matching multiple project progress curves, and adopts a limited window to limit the dynamic time warping path model for matching multiple project progress curves to solve the project construction progress curve. and The offset range of the matching points, such as Figure 4 As shown in Figure 1, the dynamic time warping path matched by multiple project schedule curves cannot enter the gray boundary area. This method is called a massive project schedule curve pattern recognition method based on dynamic timing curve fusion. The specific steps are as follows: Figure 5 shown.

[0165] S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ;

[0166] S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is equal in length to the second dynamic time warping path;

[0167] The expression of the comprehensive project progress curve sequence is as follows:

[0168] ;

[0169] in: and Represent the projects to be identified and the project progress curve respectively and The degree of membership of the progress curve of the corresponding project; Represents the best matching set of the project to be identified and the project schedule characteristics Sum the membership of the progress curves of all projects in;

[0170] S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; The expression is as follows:

[0171] ;

[0172] in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point positions A collection of; express The number of elements in the collection; Indicates the first term curve;

[0173] S45, determining whether the maximum number of iterations is met;

[0174] If satisfied, the final project progress curve model is output;

[0175] If not satisfied, the project progress curve Updated to , and repeat steps S41-S45 until the The best matching set of project progress timing characteristics corresponding to the projects to be identified All the project progress curves in the project are integrated to obtain the final project progress curve model.

[0176] In this embodiment, the progress curve pattern of the 110kV power transmission and transformation project is analyzed based on the progress time series feature matching strategy and the massive project progress curve pattern recognition method, and the recognition results obtained are as follows: Figure 6-Figure 10 shown.

[0177] In this embodiment, the K-Means clustering algorithm and the deep embedded clustering algorithm are used to compare with this solution to verify the effectiveness of this solution. Mean square error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), accuracy (ACC), and determination coefficient (R2) are selected as recognition result evaluation indicators, and the comprehensive performance of each recognition algorithm is described in detail, as shown in the following table:

[0178]

[0179] By comparing the performance indicators of the three algorithms, it is found that the determination coefficient R 2 The proposed method is 0.981, which is 8.97% and 3.42% higher than the K-means clustering algorithm and the deep embedded clustering algorithm, respectively. This shows that the proposed method can best fit the nonlinear relationship between the timing characteristics of power grid infrastructure projects and the project progress curve pattern. Compared with the K-Means clustering algorithm and the deep embedded clustering algorithm, the accuracy of the proposed method in massive project progress curve pattern recognition is increased by 6.81% and 4.75%, respectively, which further verifies the advanced nature of the proposed method in power grid project progress curve pattern recognition. For the problem of massive power grid project progress curve pattern recognition, the proposed method has the smallest mean square error and mean absolute error, which is about 3.87 lower than the K-Means clustering algorithm and about 1.23 lower than the deep embedded clustering algorithm, indicating that the proposed algorithm has the best stability and adaptability. Overall, the algorithm of this scheme has higher accuracy and effectiveness than other algorithms.

[0180] Embodiment 2:

[0181] See also Fig.11 , a power grid project planning pattern recognition system based on time series feature matching, the system comprising:

[0182] The project progress time series feature acquisition module 1 is used to construct a historical project progress curve pattern library, obtain the project progress time series features, and allocate the weight of each project progress time series feature according to the discrete degree;

[0183] Furthermore, the project progress time series feature acquisition module 1 acquires the project progress time series features and their weights according to the following steps:

[0184] S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library;

[0185] S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics;

[0186] S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows:

[0187] ;

[0188] in: Indicates The weights of the project schedule timing characteristics, Represents the first The sample variance of the project schedule timing characteristics.

[0189] The project progress timing characteristics best matching set construction module 2 is used to construct a project progress curve membership function model based on the weight of the project progress timing characteristics, and match the project to be identified with the project progress timing characteristics in the project progress curve membership function model to obtain the best matching set of project progress timing characteristics;

[0190] Furthermore, the project progress timing feature best matching set construction module 2 obtains the project progress timing feature best matching set according to the following steps:

[0191] S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance;

[0192] S22, inputting the project to be identified into the project progress curve membership function model to match the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics;

[0193] The step S22 specifically includes:

[0194] S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified;

[0195] S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them;

[0196] S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model;

[0197] The expression of the project progress curve membership function model is as follows:

[0198] ;

[0199] in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project;

[0200] S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223;

[0201] S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

[0202] A dynamic time warping path model building module 3 is used to define a distance matrix of a project schedule curve based on the best matching set of project schedule timing characteristics, and to define a first dynamic time warping path and conditions based on the distance matrix to obtain a dynamic time warping path model;

[0203] Furthermore, the dynamic time warping path model construction module 3 constructs a dynamic time warping path model according to the following steps:

[0204] S31. Assume that the project progress curve with the highest matching degree in the best matching set of project progress timing characteristics is , The sequence length is ; Assume that the progress curve of any other project in the best matching set of project progress timing characteristics is , The sequence length is ; Define the project progress curve and The distance matrix is ​​as follows:

[0205] ;

[0206] in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves;

[0207] S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows:

[0208] ;

[0209] in: represents the starting point of the first dynamic time warping path; represents the end point of the first dynamic time warping path; represents the length of the first dynamic time warping path, which satisfies , ; represents the first dynamic time warping path No. Project progress curve at each location sequence points and sequence points Match;

[0210] S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows:

[0211] ;

[0212] in: represents the first dynamic time warping path Middle Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the first dynamic time warping path Two adjacent points in the .

[0213] The project progress curve pattern acquisition module 4 is used to solve the dynamic time warping path model, obtain the second dynamic time warping path, and expand the project progress curve according to the second dynamic time warping path to obtain a new project progress curve sequence, and calculate the comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, all project progress curves are merged with all project progress timing features in the best matching set of project progress timing features to obtain the final project progress curve pattern.

[0214] Furthermore, the project progress curve mode acquisition module 4 acquires the final project progress curve mode according to the following steps:

[0215] S41. Solve the first dynamic time warping path model based on the constrained dynamic time warping algorithm to obtain a project progress curve and The corresponding second dynamic time warping path ;

[0216] S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ;

[0217] S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is equal in length to the second dynamic time warping path;

[0218] The expression of the comprehensive project progress curve sequence is as follows:

[0219] ;

[0220] in: and Represent the projects to be identified and the project progress curve respectively and The degree of membership of the progress curve of the corresponding project; Represents the best matching set of the project to be identified and the project schedule characteristics Sum the membership of the progress curves of all projects in;

[0221] S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; The expression is as follows:

[0222] ;

[0223] in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point locations A collection of; express The number of elements in the collection; Indicates the first term curve;

[0224] S45, determining whether the maximum number of iterations is met;

[0225] If satisfied, the final project progress curve model is output;

[0226] If not satisfied, the project progress curve Updated to , and repeat steps S41-S45 until the The best matching set of project progress timing characteristics corresponding to the projects to be identified All the project progress curves in the project are integrated to obtain the final project progress curve model.

[0227] Embodiment 3:

[0228] See also Fig.12 , In this embodiment, a power grid project planning pattern recognition device based on time series feature matching is also included, and the device includes a processor 5 and a memory 6;

[0229] The memory 6 is used to store the computer program code 61 and transmit the computer program code 61 to the processor 5;

[0230] The processor 5 is used to execute the power grid project planning pattern recognition method based on time series feature matching described in Example 1 according to the instructions in the computer program code 61.

[0231] This embodiment also includes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed on a computer, the power grid project planning pattern recognition method based on time series feature matching described in Example 1 is implemented.

[0232] Generally speaking, the computer instructions for implementing the method of the present invention may be carried in any combination of one or more computer-readable storage media. Non-transitory computer-readable storage media may include any computer-readable media, except for the signal itself that is temporarily propagating.

[0233] Computer readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or components, or any combination thereof. More specific examples (non-exhaustive list) of computer readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EKROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or device.

[0234] Computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, SMalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages, in particular, Python suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. can be used. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer or to an external computer (for example, using an Internet service provider to connect via the Internet) through any type of network, including a local area network (LAN) or a wide area network (WAN).

[0235] The above-mentioned device and non-temporary computer-readable storage medium can refer to the specific description of a power grid project planning pattern recognition method based on time series feature matching and its beneficial effects, which will not be repeated here.

[0236] Although the embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A power grid project planning pattern recognition method based on time series feature matching, characterized in that: include: S1. Build a historical project progress curve pattern library, obtain project progress timing characteristics, and assign weights to each project progress timing characteristics according to the degree of discreteness; S2. Based on the weights of the project progress timing characteristics, a project progress curve membership function model is constructed, and the project to be identified is matched with the project progress timing characteristics in the project progress curve membership function model to obtain the best matching set of project progress timing characteristics; S3. Based on the best matching set of project progress timing characteristics, define a distance matrix of the project progress curve, and based on the distance matrix, define a first dynamic time warping path and conditions to obtain a dynamic time warping path model; S4. Solve the dynamic time warping path model to obtain a second dynamic time warping path, expand the project progress curve according to the second dynamic time warping path, obtain a new project progress curve sequence, and calculate a comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, merge all project progress curves with all project progress timing features in the best matching set of project progress timing features to obtain a final project progress curve pattern; The step S3 specifically includes: S31. Assume that the project progress curve with the highest matching degree in the best matching set is ; Assume that the progress curve of any other project in the best matching set is ; Define the project progress curve and The distance matrix is ​​as follows: ； in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves, , Respectively , The length of the sequence; S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows: ； in: , denote the starting point and the end point of the first dynamic time warping path respectively; represents the length of the first dynamic time warping path; S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows: ； in: represents the first dynamic time warping path Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the satisfying condition of two adjacent points in the first dynamic time warping path; The step S4 specifically includes: S41. Based on the constrained dynamic time warping algorithm, solve the dynamic time warping path model and obtain and The corresponding second dynamic time warping path ; S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ; S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is the same length as the second dynamic time warping path; its expression is as follows: ； in: and Represent the projects to be identified and the project progress curve respectively , The degree of membership of the progress curve of the corresponding project; It represents the sum of the membership degrees of the progress curves of all projects in the best matching set between the project to be identified and the project progress timing characteristics; S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; Its expression is as follows: ； in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point locations A collection of; express The number of elements in the collection; Indicates the first term curve; S45, determine whether the maximum number of iterations is met; if so, output the final project progress curve model; if not, Updated to , and repeat steps S41-S45 until the All the project progress curves in the best matching set of project progress timing characteristics corresponding to the projects to be identified are merged to obtain the final project progress curve pattern.

2. The power grid project planning pattern recognition method based on time series feature matching according to claim 1 is characterized in that: The step S1 specifically includes: S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library; S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics; S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows: ； in: Indicates The weights of the project schedule timing characteristics, Represents the historical project progress curve pattern library The sample variance of the project schedule timing characteristics.

3. The method for identifying a pattern in power grid project planning based on time series feature matching according to claim 2 is characterized in that: The step S2 specifically includes: S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance; S22, inputting the project to be identified into the project progress curve membership function model to match it with the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics.

4. The method for identifying a pattern in power grid project planning based on time series feature matching according to claim 3 is characterized in that: The step S22 specifically includes: S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified; S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them; S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model; The expression of the project progress curve membership function model is as follows: ； in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project; S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223; S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

5. A power grid project planning pattern recognition system based on time series feature matching, characterized in that: The system comprises: The project progress timing feature acquisition module is used to build a historical project progress curve pattern library, obtain the project progress timing features, and assign the weight of each project progress timing feature according to the discrete degree; A project progress timing feature best matching set building module is used to build a project progress curve membership function model based on the weight of the project progress timing feature, and match the project to be identified with the project progress timing feature in the project progress curve membership function model to obtain the best matching set of project progress timing features; A dynamic time warping path model building module is used to define a distance matrix of a project schedule curve based on the best matching set of project schedule timing characteristics, and to define a first dynamic time warping path and conditions based on the distance matrix to obtain a dynamic time warping path model; The project progress curve pattern acquisition module is used to solve the dynamic time warping path model, obtain the second dynamic time warping path, and expand the project progress curve according to the second dynamic time warping path to obtain a new project progress curve sequence, and calculate the comprehensive project progress curve sequence; based on the comprehensive project progress curve sequence, all project progress curves are merged with all project progress timing features in the best matching set of project progress timing features to obtain the final project progress curve pattern; The dynamic time warping path model construction module constructs a dynamic time warping path model according to the following steps: S31. Assume that the project progress curve with the highest matching degree in the best matching set is ; Assume that the progress curve of any other project in the best matching set is ; Define the project progress curve and The distance matrix is ​​as follows: ； in: represents the distance matrix, express Middle Item and Middle The distance between the corresponding project progress curves, , Respectively , The length of the sequence; S32. Based on the distance matrix, define the first dynamic time warping path for matching the progress curves of multiple projects as follows: ； in: , denote the starting point and the end point of the first dynamic time warping path respectively; represents the length of the first dynamic time warping path; S33, setting the conditions that the first dynamic time warping path needs to meet, and constructing a dynamic time warping path model for matching multi-project progress curves; the expression of the dynamic time warping path model is as follows: ； in: represents the first dynamic time warping path Matching points Project progress curve and The distance between the corresponding construction progress completion rates, and represents the boundary conditions, and represents the satisfying condition of two adjacent points in the first dynamic time warping path; The project progress curve mode acquisition module acquires the final project progress curve mode according to the following steps: S41. Based on the constrained dynamic time warping algorithm, solve the dynamic time warping path model and obtain and The corresponding second dynamic time warping path ; S42. Project progress curve and Expand into a new project progress curve sequence according to the second dynamic time warping path and ; S43, based on the new project progress curve sequence and , calculate and obtain the comprehensive project progress curve sequence ; is the same length as the second dynamic time warping path; its expression is as follows: ； in: and Represent the projects to be identified and the project progress curve respectively , The degree of membership of the progress curve of the corresponding project; It represents the sum of the membership degrees of the progress curves of all projects in the best matching set between the project to be identified and the project progress timing characteristics; S44. Project progress curve Integration In The fusion sequence of project progress curves corresponding to the projects to be identified ; Its expression is as follows: ； in: Indicates that the second dynamic time warping path satisfies Project progress curve and All matching point positions A collection of; express The number of elements in the collection; Indicates the first term curve; S45, determine whether the maximum number of iterations is met; if so, output the final project progress curve model; if not, Updated to , and repeat steps S41-S45 until the All the project progress curves in the best matching set of project progress timing characteristics corresponding to the projects to be identified are merged to obtain the final project progress curve pattern.

6. A power grid project planning pattern recognition system based on time series feature matching according to claim 5, characterized in that: The project progress timing feature acquisition module acquires the project progress timing features and their weights according to the following steps: S11. Obtain historical project progress curve patterns of power grid infrastructure projects and construct a historical project progress curve pattern library; S12. Filter and match the required project progress timing characteristics from the historical project progress curve pattern library and normalize them to obtain the project progress timing characteristics; S13, allocating the weight of each project progress time series feature according to the discrete degree of all historical project progress curve data in the historical project progress curve pattern library; the weight expression is as follows: ； in: Indicates The weights of the project schedule timing characteristics, Represents the historical project progress curve pattern library The sample variance of the project schedule timing characteristics.

7. A power grid project planning pattern recognition system based on time series feature matching according to claim 6, characterized in that: The project progress timing feature best matching set construction module obtains the project progress timing feature best matching set according to the following steps: S21. Based on the weight of the project progress time series characteristics, the project progress curve membership function model is constructed using weighted Hamming distance; S22, inputting the project to be identified into the project progress curve membership function model to match the project progress timing characteristics, and obtaining the best matching set of project progress timing characteristics; The step S22 specifically includes: S221, inputting a historical project progress curve pattern library and a project progress time series characteristic value of the project to be identified; S222, extracting the project progress time series characteristic values ​​of all historical projects in the historical project progress curve pattern library, and normalizing them; S223, calculating the project progress curve membership of the to-be-identified project and the historical project based on the project progress curve membership function model; The expression of the project progress curve membership function model is as follows: ； in: Indicates the item to be identified. The project progress time series characteristic value, Represents the first item to be identified after normalization The project progress time series characteristic value, Represents the normalized historical project progress curve pattern library History Project The project progress time series characteristic value, Indicates the progress curve pattern library of the project to be identified and the historical project The degree of membership of the historical project; S224, judging whether the project progress curve membership of the project to be identified matches that of the historical project; if so, outputting the historical project number in the corresponding historical project progress curve pattern library; if not, returning to step S223; S225. Repeat steps S221-S224 until all historical projects have been retrieved and the project progress curve memberships of all projects to be identified have been calculated, and then output the best matching set of project progress timing characteristics.

Images