A multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combination weighting and cosine similarity improved TOPSIS
By constructing a multidimensional risk assessment index system and a combined weighting mechanism, and by improving the TOPSIS model with cosine similarity, the problem of multidimensional risk assessment for compressed air energy storage projects has been solved, achieving accurate risk assessment and decision support throughout the entire life cycle.
Patent Information
- Application Number
- CN202610348229.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-20
- Publication Date
- 2026-08-25
AI Technical Summary
Existing risk assessment methods for compressed air energy storage projects suffer from incomplete indicator coverage, subjective or insufficiently objective indicator weighting methods, and poor adaptability of traditional TOPSIS models, making it impossible to conduct comprehensive and accurate multidimensional risk assessments.
A multi-dimensional risk assessment index system is constructed, and a subjective and objective weighting mechanism combining the analytic hierarchy process and the CRITIC method is adopted. The TOPSIS model is improved by combining cosine similarity. The location and shape differences of the evaluation objects are characterized by Euclidean distance and cosine similarity, and the comprehensive proximity is calculated to realize the risk assessment of the entire life cycle of compressed air energy storage projects.
It enables comprehensive, accurate, and quantifiable risk assessment of compressed air energy storage projects throughout their entire lifecycle, improving the robustness and accuracy of the assessment results, adapting to risk assessments of different scales and conditions, and providing reliable decision support.
Smart Images

Figure CN122636103A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of energy engineering and power system operation risk assessment technology, specifically to a multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS. Background Technology
[0002] With the rapid growth of installed capacity of new energy sources, the intermittency, volatility, and anti-peak-shaving characteristics of wind and photovoltaic power generation are becoming increasingly prominent, posing severe challenges to the safe and stable operation of the power system and the capacity for new energy absorption. Against this backdrop, large-scale energy storage technology has become a core technological means to improve the power system's regulation capabilities, smooth out fluctuations in new energy output, and alleviate the pressure on new energy absorption.
[0003] Compressed air energy storage (CAES) technology boasts significant advantages such as large storage capacity, long cycle life, stable operating efficiency, controllable construction costs, and strong geological adaptability, making it a crucial development direction for new energy storage technologies in modern power systems. Currently, numerous demonstration power plants of various types of CAES have been built both domestically and internationally, while a large number of new projects are planned. These projects exhibit diversified development trends in terms of engineering geological conditions, energy storage technology pathways, system integration schemes, and business models.
[0004] Compressed air energy storage projects are characterized by long construction periods, large investment scales, high technological integration, long operation cycles, and significant susceptibility to policy and market conditions. Throughout their entire lifecycle, they face multi-dimensional risk factors related to policy, economy, market, technology, natural environment, and operation management. However, current risk assessment methods for compressed air energy storage projects still suffer from the following core shortcomings: The indicator system is incomplete, mostly focusing only on a single dimension of risk, and cannot systematically and structurally depict the multidimensional risks throughout the entire project lifecycle, resulting in blind spots in risk identification; There are limitations to the methods of assigning weights to indicators. The single subjective weighting method relies too much on expert experience and has a strong degree of subjectivity and arbitrariness; the single objective weighting method relies only on the statistical characteristics of the data itself and cannot incorporate the engineering experience and domain knowledge of industry experts, resulting in a deviation between the weighting results and the actual engineering situation. Traditional TOPSIS evaluation models only use Euclidean distance to characterize the positional difference between the evaluation object and the ideal solution, but cannot characterize the shape difference between the evaluation vector and the ideal solution. They are poorly adaptable to heterogeneous indicators, extreme data, and differences in indicator distribution, and are prone to problems such as deviation between evaluation results and actual conditions and insufficient robustness of ranking results. They cannot provide reliable quantitative support for project planning, site selection, scheme demonstration, and investment decisions. Summary of the Invention
[0005] The purpose of this invention is to provide a multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-based improved TOPSIS, so as to achieve a comprehensive, accurate and quantifiable evaluation of the risks throughout the entire life cycle of compressed air energy storage projects.
[0006] To achieve the above objectives, the present invention employs the following technical solution: 1. A multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-based improved TOPSIS, characterized by the following steps: S1. Construct a multi-dimensional risk assessment indicator system covering the entire life cycle of compressed air energy storage projects; S2. Subjective weighting is used to calculate the subjective weight of each evaluation indicator in the risk assessment indicator system, and objective weighting is used to calculate the objective weight of each evaluation indicator. S3. Based on the preset objective preference coefficient, the subjective weight and the objective weight are linearly fused to obtain the combined weight of each evaluation index; S4. Obtain the original data of each evaluation index for at least two compressed air energy storage projects to be evaluated, perform dimensionless processing on the original data, and construct a weighted standardized decision matrix by combining the combined weights. S5. Based on the weighted standardized decision matrix, determine the positive ideal solution and the negative ideal solution for risk assessment, and calculate the Euclidean distance and cosine similarity between each project to be evaluated and the positive ideal solution and the negative ideal solution, respectively. S6. Normalize the Euclidean distance and the cosine similarity respectively, and calculate the comprehensive closeness of each item to be evaluated to the positive ideal solution and the negative ideal solution based on the preset position preference coefficient and shape preference coefficient. S7. Calculate the relative proximity of each compressed air energy storage project to be evaluated based on the comprehensive proximity degree, and rank the comprehensive risk level of each compressed air energy storage project according to the relative proximity degree. The larger the value of the relative proximity degree, the lower the comprehensive risk of the corresponding compressed air energy storage project.
[0007] 2. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS according to claim 1, characterized in that, in step S2, the subjective weighting is the analytic hierarchy process, and the calculation of the subjective weights specifically includes the following steps: S211. Based on the aforementioned risk assessment index system, construct a judgment matrix A for pairwise index comparison: , In the formula, Let be the importance scale value of indicator j relative to indicator i, and n be the number of evaluation indicators participating in the comparison. Matrix A satisfies , , ; S212. Calculate the largest eigenvalue of the judgment matrix A. And calculate the consistency test index CI: Query the average random consistency index RI for the corresponding number of indicators n, and calculate the consistency ratio CR: The consistency test is determined based on the relationship between CR and the preset consistency threshold. If the consistency test is not passed, the judgment matrix A is reconstructed until the consistency test is passed. S213. Standardize the judgment matrix A that passes the consistency test to construct a standardized matrix: , , The subjective weights of each evaluation indicator were calculated. : , , Where j = 1, 2, ..., n.
[0008] 3. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS according to claim 1, characterized in that, in step S2, the objective weighting is the CRITIC method, and the calculation of the objective weights specifically includes the following steps: S221. The original data matrix for m items to be evaluated and n evaluation indicators. Perform positive transformation: , Obtain the normalized matrix ; in, Let be the original score for the j-th evaluation indicator of the i-th item to be evaluated. Let i = 1, 2, ..., m, j = 1, 2, ..., n, representing the score of the j-th evaluation indicator for the i-th item to be evaluated after positive transformation. S222. Calculate the standard deviation of the j-th evaluation index. : , In the formula, Let be the mean of all positively oriented scores for the j-th evaluation indicator; S223. Calculate the correlation coefficient between the j-th evaluation index and the k-th evaluation index. : , In the formula, Let be the mean of all positively oriented scores for the k-th evaluation indicator, where k = 1, 2, ..., n; S224. Calculate the information content of the j-th evaluation index. : ; S225. Calculate the objective weight of the j-th evaluation indicator. : .
[0009] 4. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS according to claim 1, characterized in that, in step S3, the calculation formula for the combined weights is: , In the formula, Let be the combined weight of the j-th evaluation index, and u be the objective preference coefficient, with a value range of 0 ≤ u ≤ 1; Let be the objective weight of the j-th evaluation indicator; Let be the subjective weight of the j-th evaluation indicator.
[0010] 5. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 3, characterized in that step S4 specifically includes the following steps: S41, According to the formula The original data after normalization is dimensionless to obtain a standardized matrix. ; S42. Construct a weighted standardized decision matrix by combining the aforementioned weights. The calculation formula is: In the formula, Let j be the weighted standardized value of the j-th evaluation indicator for the i-th item to be evaluated. is the combined weight of the j-th evaluation index.
[0011] 6. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that step S5 specifically includes the following steps: S51. Determining the positive ideal solution Y based on the weighted standardized decision matrix. + With negative ideal solution Y - : , , In the formula, Let be the positive ideal solution value corresponding to the j-th evaluation index. Let be the negative ideal solution value corresponding to the j-th evaluation index; S52. Calculate the Euclidean distance between the i-th item to be evaluated and the positive ideal solution. Euclidean distance to the negative ideal solution : , ; S53 Calculate the cosine similarity between the i-th item to be evaluated and the positive ideal solution. Cosine similarity with negative ideal solutions : , .
[0012] 7. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 6, characterized in that the normalization process in step S6 specifically includes the following steps: S61. Normalize the Euclidean distance to obtain the dimensionless Euclidean distance. , : , , In the formula, For all items to be evaluated The maximum value, The minimum value of $D^+_i$ for all items to be evaluated; For all items to be evaluated The maximum value, For all items to be evaluated The minimum value; S62. Normalize the cosine similarity to obtain the dimensionless cosine similarity. , : , , In the formula, For all items to be evaluated The maximum value, For all items to be evaluated The minimum value; For all items to be evaluated The maximum value, For all items to be evaluated The minimum value.
[0013] 8. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 7, characterized in that the calculation of the comprehensive similarity specifically includes the following steps: The degree of comprehensive similarity between the i-th item to be evaluated and the positive ideal solution The degree of comprehensive proximity to the negative ideal solution : , , In the formula, This is the location preference coefficient. Let be the shape preference coefficient, satisfying ,and , .
[0014] 9. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity improved TOPSIS according to claim 1, characterized in that, in step S7, the relative proximity... The calculation formula is: , In the formula, Let be the overall similarity between the i-th item to be evaluated and the positive ideal solution. The comprehensive similarity between the i-th item to be evaluated and the negative ideal solution.
[0015] 10. The method for comprehensive evaluation of multidimensional risks of compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS according to claim 1, characterized in that, in step S1, the multidimensional risk evaluation index system includes a first-level risk index, and each first-level risk index is provided with a corresponding second-level risk index. The primary risk indicators are policy risk, economic risk, market risk, technological risk, natural risk, and operational risk. The secondary risks include: policy-related risks such as policy support stability risk, regulatory and approval risk, and market system change risk; economic-related risks such as investment cost control risk, financing and interest rate risk, and return uncertainty risk; market-related risks such as electricity price and demand fluctuation risk, competition and substitution risk, and market access and transaction risk; technology-related risks such as technology maturity risk, system integration and compatibility risk, and equipment reliability and performance risk; natural-related risks such as geological and gas storage safety risk, and climate and environmental adaptability risk; and operational risks such as safety production risk, operation and maintenance management level risk, control and dispatch risk, and supply chain and spare parts guarantee risk.
[0016] The advantages of this invention are: This invention constructs a risk assessment indicator system covering six dimensions: policy environment, economic input, market competition, technological reliability, natural conditions, and operation management. It includes 18 secondary sub-indicators, which can comprehensively and structurally characterize the key risk factors throughout the entire life cycle of compressed air energy storage projects, achieving blind-spot-free coverage of risk identification and solving the problem of incomplete coverage in existing technical indicator systems.
[0017] This invention employs a subjective and objective weighting mechanism that combines the Analytic Hierarchy Process (AHP) with the CRITIC method. The AHP incorporates the engineering experience and industry knowledge of energy experts, while the CRITIC method fully explores the variability and conflict of the indicator data itself. This overcomes the shortcomings of traditional single weighting methods, which are either too subjective or lack objectivity. As a result, the allocation of indicator weights is more reasonable and stable, and more adaptable to actual engineering scenarios.
[0018] This invention improves the traditional TOPSIS model by introducing cosine similarity. It uses Euclidean distance to represent the positional difference between the evaluation object and the ideal solution, and cosine similarity to represent the shape difference between the evaluation object and the ideal solution. This solves the problems of the traditional TOPSIS model, which only considers positional differences and has poor adaptability to heterogeneous indicators and extreme data, and significantly improves the robustness and accuracy of the evaluation results.
[0019] This invention achieves flexible weight allocation for location and shape differences through location preference coefficients and shape preference coefficients, and achieves flexible integration of subjective and objective weights through objective preference coefficients. It can be adapted to risk assessment scenarios of compressed air energy storage projects of different scales, technical paths, geological conditions, and business models. The model has strong universality and the calculation steps are standardized and clear. It can provide quantifiable and comparable scientific and technological support for the planning, site selection, scheme demonstration, and investment decision-making of compressed air energy storage projects. Attached Figure Description
[0020] Figure 1 This is a flowchart of the method described in this invention; Figure 2 This is a diagram of the CAES project risk assessment index system of the present invention. Detailed Implementation
[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
[0022] This embodiment provides a multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS, including the following steps: S1. Construct a multi-dimensional risk assessment indicator system covering the entire life cycle of compressed air energy storage projects; S2. Subjective weighting is used to calculate the subjective weight of each evaluation indicator in the risk assessment indicator system, and objective weighting is used to calculate the objective weight of each evaluation indicator. S3. Based on the preset objective preference coefficient, the subjective weight and the objective weight are linearly fused to obtain the combined weight of each evaluation index; S4. Obtain the original data of each evaluation index for at least two compressed air energy storage projects to be evaluated, perform dimensionless processing on the original data, and construct a weighted standardized decision matrix by combining the combined weights. S5. Based on the weighted standardized decision matrix, determine the positive ideal solution and the negative ideal solution for risk assessment, and calculate the Euclidean distance and cosine similarity between each project to be evaluated and the positive ideal solution and the negative ideal solution, respectively. S6. Normalize the Euclidean distance and the cosine similarity respectively, and calculate the comprehensive closeness of each item to be evaluated to the positive ideal solution and the negative ideal solution based on the preset position preference coefficient and shape preference coefficient. S7. Calculate the relative proximity of each compressed air energy storage project to be evaluated based on the comprehensive proximity degree, and rank the comprehensive risk level of each compressed air energy storage project according to the relative proximity degree. The larger the value of the relative proximity degree, the lower the comprehensive risk of the corresponding compressed air energy storage project.
[0023] The following section will use 10 typical compressed air energy storage projects from both domestic and international sources as examples to illustrate the above steps. The specific process is as follows: Figure 1 As shown, the projects to be evaluated include the Jintan Salt Cavern Compressed Air Energy Storage Project in Jiangsu, China; the Zhangjiakou Advanced Compressed Air Energy Storage Project in Hebei, China; the Yingcheng Advanced Compressed Air Energy Storage Project in Hubei, China; the Feicheng Compressed Air Energy Storage Project in Shandong, China; the Xinyang Artificial Chamber Compressed Air Energy Storage Project in Henan, China; the Hantford Compressed Air Energy Storage Project in Germany; the Alabama Compressed Air Energy Storage Project in the United States; the Manchester Liquid Air Energy Storage Project in the United Kingdom; the California Insulated Compressed Air Energy Storage Project in the United States; and the New Hampshire Thermal Storage Compressed Air Energy Storage Project in the United States.
[0024] Please refer to the multidimensional risk assessment indicator system. Figure 2It includes 6 primary risk indicators and 18 secondary risk indicators, specifically: Primary indicators: Policy risk A1, Economic risk A2, Market risk A3, Technological risk A4, Natural risk A5, Operational risk A6; Secondary indicators: A11 Policy support stability risk, A12 Regulatory and approval risk, A13 Market system change risk; A21 Investment cost control risk, A22 Financing and interest rate risk, A23 Return uncertainty risk; A31 Electricity price and demand fluctuation risk, A32 Competition and substitution risk, A33 Market access and transaction risk; A41 Technology maturity risk, A42 System integration and compatibility risk, A43 Equipment reliability and performance risk; A51 Geological and gas storage safety risk, A52 Climate and environmental adaptability risk; A61 Safe production risk, A62 Operation and maintenance management level risk, A63 Control and dispatch risk, A64 Supply chain and spare parts guarantee risk.
[0025] In step S2, the subjective weights are calculated using the analytic hierarchy process (AHP): S211. Construct the judgment matrix for each level of indicators, assign importance values to each pair of indicators, and construct judgment matrix A: , In the formula, Let be the importance scale value of indicator j relative to indicator i, and n be the number of evaluation indicators participating in the comparison. Matrix A satisfies , , .
[0026] S212. Calculate the largest eigenvalue of the judgment matrix A. And calculate the consistency test index CI: , Query the average random consistency index RI for the corresponding number of indicators n, and calculate the consistency ratio CR: The consistency test is determined based on the relationship between CR and the preset consistency threshold. If the consistency ratio CR of all judgment matrices is less than 0.1, the consistency test is passed; otherwise, the judgment matrix A is reconstructed until the consistency test is passed.
[0027] S213. Standardize the judgment matrix A that passes the consistency test to construct a standardized matrix: , , The subjective weights of each evaluation indicator were calculated. : , , Where j = 1, 2, ..., n.
[0028] The judgment matrix is standardized to calculate the subjective weights of each secondary indicator. The values are: (0.09, 0.07, 0.07, 0.07, 0.08, 0.06, 0.06, 0.06, 0.06, 0.08, 0.06, 0.04, 0.06, 0.06, 0.06, 0.03, 0.03, 0.02, 0.03).
[0029] The objective weights are calculated using the CRITIC method: S221. Obtain the original scoring data of 18 secondary indicators for 10 items to be evaluated, and construct the original data matrix. Then, perform a positive transformation to obtain the positive matrix R; calculate the standard deviation of each indicator. Correlation coefficients between indicators Then calculate the information content of each indicator. Finally, the objective weights of each secondary indicator are obtained. for: (0.13, 0.12, 0.07, 0.03, 0.03, 0.03, 0.04, 0.03, 0.05, 0.05, 0.05, 0.03, 0.11, 0.05, 0.04, 0.04, 0.04, 0.04).
[0030] In step S3, the objective preference coefficient u=0.5 is set, and the combined weights of each indicator are calculated using a linear fusion formula: , Finally, the combined weights are obtained. for: = (0.11, 0.09, 0.07, 0.05, 0.05, 0.04, 0.05, 0.05, 0.06, 0.06, 0.06, 0.04, 0.09, 0.05, 0.03, 0.03, 0.03, 0.04).
[0031] In step S4, according to the formula: , , The raw data from the 10 projects were dimensionless to obtain a standardized matrix. ; Construct a weighted standardized decision matrix by combining the combined weights The calculation formula is: In the formula, Let j be the weighted standardized value of the j-th evaluation indicator for the i-th item to be evaluated. is the combined weight of the j-th evaluation index.
[0032] In step S5, the positive ideal solution Y is determined based on the weighted standardized decision matrix. + With negative ideal solution Y - : , , In the formula, Let be the positive ideal solution value corresponding to the j-th evaluation index. Y is calculated as the negative ideal solution value corresponding to the j-th evaluation index. + = (0.037299145, 0.031365993, 0.023228659, 0.018985179, 0.018774189, 0.015850405, 0.016478654, 0.01599216, 0.018535323, Y - = (0.030106651, 0.025141773, 0.018974905, 0.013483354, 0.013903958, 0.011309555, 0.013767462, 0.012903578, 0.014922044, 0.017409462, 0.015650456, 0.009694332, 0.022515763, 0.01466833, 0.008687772, 0.008657854, 0.009282361, 0.00956098).
[0033] According to the formula: , , Calculate the Euclidean distance between the i-th item to be evaluated and the positive ideal solution. Euclidean distance to the negative ideal solution : = (0.004477965, 0.011392775, 0.006791599, 0.005249998, 0.011564504, 0.004919137, 0.007062126, 0.016221979, 0.01577223, 0.01942573). = (0.017401945, 0.011126797, 0.014931643, 0.016934772, 0.010045566, 0.018271993, 0.016470976, 0.005345871, 0.005450012, 0.002317799).
[0034] According to the formula: , , Calculate the cosine similarity between the i-th item to be evaluated and the positive ideal solution. Cosine similarity with negative ideal solutions : = (0.043383242, 0.041543039, 0.042736056, 0.043243595, 0.041331184, 0.043339816, 0.042716735, 0.039838467, 0.039916719, 0.038721182). = (0.037958476, 0.036610578, 0.03749053, 0.037860263, 0.036431331, 0.037817386, 0.037320212, 0.035264315, 0.035304014, 0.034358782).
[0035] In step S6, according to the formula: , , Normalizing the Euclidean distance yields the dimensionless Euclidean distance. , : =(0,0.4626,0.1548,0.0516,0.4741,0.0295,0.1729,0.7857,0.7556,1), =(0.9455,0.5521,0.7906,0.9162,0.4844,1,0.8871,0.1898,0.1963,0).
[0036] According to the formula: , , Normalize the cosine similarity to obtain the dimensionless cosine similarity. , : = (1, 0.605281, 0.86118018, 0.970046056, 0.559838607, 0.990685106, 0.857035919, 0.23965468, 0.256439627, 0), = (1, 0.625551943, 0.870003959, 0.972716076, 0.575757009, 0.960804987, 0.822689218, 0.251558447, 0.262586864, 0).
[0037] Set the position preference coefficient e1 = 0.5 and the shape preference coefficient e2 = 0.5, according to the formula: , , Calculate the overall similarity: = (0.972732941, 0.578712048, 0.825904435, 0.943114872, 0.522105364, 0.995342553, 0.872074586, 0.214726276, 0.226382503, 0), = (0.5, 0.544075107, 0.5123926, 0.512182391, 0.524921927, 0.495159618, 0.497784324, 0.518614342, 0.509084525, 0.5).
[0038] In step S7, the formula is used: , Calculate the relative similarity of 10 items They are respectively: = (0.660495134, 0.515424536, 0.617130886, 0.648056514, 0.498654972, 0.667790073, 0.636616355, 0.29280565, 0.30780782, 0).
[0039] Based on the relative proximity scores from highest to lowest, the comprehensive risk levels of the 10 projects, from lowest to highest, are as follows: Hantford Compressed Air Energy Storage Project (Germany), Jintan Salt Cavern Compressed Air Energy Storage Project (Jiangsu, China), Feicheng Compressed Air Energy Storage Project (Shandong, China), Compressed Air Energy Storage Project (Alabama, USA), Advanced Compressed Air Energy Storage Project (Yingcheng, Hubei, China), Advanced Compressed Air Energy Storage Project (Zhangjiakou, Hebei, China), Artificial Chamber Compressed Air Energy Storage Project (Xinyang, Henan, China), Thermal Compressed Air Energy Storage Project (California, USA), Liquid Air Energy Storage Project (Manchester, UK), and Thermal Storage Compressed Air Energy Storage Project (New Hampshire, USA).
[0040] The ranking results of the present invention compared with the traditional TOPSIS method that only uses Euclidean distance are shown in the table below.
[0041] Table 1. Comparison of results between the method of this invention and the traditional TOPSIS method. The comparison shows that the traditional TOPSIS method only considers the positional differences in Euclidean distance and cannot identify the shape differences in the evaluation vectors, resulting in a significant deviation in the ranking of the Feicheng project in Shandong, China and the Alabama project in the United States. In contrast, the present invention introduces cosine similarity to characterize both positional and shape differences, and the evaluation results are more in line with the actual risk level of the project, significantly improving the evaluation accuracy.
[0042] Example 2 This embodiment sets different preference coefficients for different situations and compares the ranking differences of the comprehensive risk levels of 10 items under different preferences.
[0043] Table 2 Comparison of absorption results with different preferences Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-based improved TOPSIS, characterized in that, Includes the following steps: S1. Construct a multi-dimensional risk assessment indicator system covering the entire life cycle of compressed air energy storage projects; S2. Subjective weighting is used to calculate the subjective weight of each evaluation indicator in the risk assessment indicator system, and objective weighting is used to calculate the objective weight of each evaluation indicator. S3. Based on the preset objective preference coefficient, the subjective weight and the objective weight are linearly fused to obtain the combined weight of each evaluation index. S4. Obtain the original data of each evaluation index for at least two compressed air energy storage projects to be evaluated, perform dimensionless processing on the original data, and construct a weighted standardized decision matrix by combining the combined weights. S5. Based on the weighted standardized decision matrix, determine the positive ideal solution and the negative ideal solution for risk assessment, and calculate the Euclidean distance and cosine similarity between each project to be evaluated and the positive ideal solution and the negative ideal solution, respectively. S6. Normalize the Euclidean distance and the cosine similarity respectively, and calculate the comprehensive closeness of each item to be evaluated to the positive ideal solution and the negative ideal solution based on the preset position preference coefficient and shape preference coefficient. S7. Calculate the relative proximity of each compressed air energy storage project to be evaluated based on the comprehensive proximity degree, and rank the comprehensive risk level of each compressed air energy storage project according to the relative proximity degree. The larger the value of the relative proximity degree, the lower the comprehensive risk of the corresponding compressed air energy storage project.
2. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that, In step S2, the subjective weighting is performed using the analytic hierarchy process (AHP), and the calculation of the subjective weights specifically includes the following steps: S211. Based on the aforementioned risk assessment index system, construct a judgment matrix A for pairwise index comparison: , In the formula, Let be the importance scale value of indicator j relative to indicator i, and n be the number of evaluation indicators participating in the comparison. Matrix A satisfies , , ; S212. Calculate the largest eigenvalue of the judgment matrix A. And calculate the consistency test index CI: Query the average random consistency index RI for the corresponding number of indicators n, and calculate the consistency ratio CR: The consistency test is determined based on the relationship between CR and the preset consistency threshold. If the consistency test is not passed, the judgment matrix A is reconstructed until the consistency test is passed. S213. Standardize the judgment matrix A that passes the consistency test to construct a standardized matrix: , , The subjective weights of each evaluation indicator were calculated. : , , Where j = 1, 2, ..., n.
3. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that, In step S2, the objective weighting is performed using the CRITIC method, and the calculation of the objective weights specifically includes the following steps: S221. The original data matrix for m items to be evaluated and n evaluation indicators. Perform positive transformation: , Obtain the normalized matrix ; in, Let be the original score for the j-th evaluation indicator of the i-th item to be evaluated. Let i = 1, 2, ..., m, j = 1, 2, ..., n; S222. Calculate the standard deviation of the j-th evaluation index. : , In the formula, Let be the mean of all positively oriented scores for the j-th evaluation indicator; S223. Calculate the correlation coefficient between the j-th evaluation index and the k-th evaluation index. : , In the formula, Let be the mean of all positively oriented scores for the k-th evaluation indicator, where k = 1, 2, ..., n; S224. Calculate the information content of the j-th evaluation index. : ; S225. Calculate the objective weight of the j-th evaluation indicator. : 。 4. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that, In step S3, the formula for calculating the combined weights is: , In the formula, Let be the combined weight of the j-th evaluation index, and u be the objective preference coefficient, with a value range of 0 ≤ u ≤ 1; Let be the objective weight of the j-th evaluation indicator; Let be the subjective weight of the j-th evaluation indicator.
5. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 3, characterized in that, Step S4 specifically includes the following steps: S41, According to the formula The original data after normalization is dimensionless to obtain a standardized matrix. ; S42. Construct a weighted standardized decision matrix by combining the aforementioned weights. The calculation formula is: In the formula, Let j be the weighted standardized value of the j-th evaluation indicator for the i-th item to be evaluated. is the combined weight of the j-th evaluation index.
6. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that, Step S5 specifically includes the following steps: S51. Determining the positive ideal solution Y based on the weighted standardized decision matrix. + With negative ideal solution Y - : , , In the formula, Let be the positive ideal solution value corresponding to the j-th evaluation index. Let be the negative ideal solution value corresponding to the j-th evaluation index; S52. Calculate the Euclidean distance between the i-th item to be evaluated and the positive ideal solution. Euclidean distance to the negative ideal solution : , ; S53 Calculate the cosine similarity between the i-th item to be evaluated and the positive ideal solution. Cosine similarity with negative ideal solutions : , 。 7. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 6, characterized in that, The normalization process in step S6 specifically includes the following steps: S61. Normalize the Euclidean distance to obtain the dimensionless Euclidean distance. , : , , In the formula, For all items to be evaluated The maximum value, The minimum value of $D^+_i$ for all items to be evaluated; For all items to be evaluated The maximum value, For all items to be evaluated The minimum value; S62. Normalize the cosine similarity to obtain the dimensionless cosine similarity. , : , , In the formula, For all items to be evaluated The maximum value, For all items to be evaluated The minimum value; For all items to be evaluated The maximum value, For all items to be evaluated The minimum value.
8. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 7, characterized in that, The calculation of overall proximity includes the following steps: The degree of comprehensive similarity between the i-th item to be evaluated and the positive ideal solution The degree of comprehensive proximity to the negative ideal solution : , , In the formula, This is the location preference coefficient. Let be the shape preference coefficient, satisfying ,and , .
9. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS as described in claim 1, characterized in that, In step S7, the relative proximity The calculation formula is: , In the formula, Let be the overall similarity between the i-th item to be evaluated and the positive ideal solution. The comprehensive similarity between the i-th item to be evaluated and the negative ideal solution.
10. The multi-dimensional risk comprehensive evaluation method for compressed air energy storage projects based on combined weighting and cosine similarity-improved TOPSIS according to claim 1, characterized in that, In step S1, the multidimensional risk assessment index system includes primary risk indicators, and each primary risk indicator is provided with a corresponding secondary risk indicator. The primary risk indicators are policy risk, economic risk, market risk, technological risk, natural risk, and operational risk. The secondary risks include: policy-related risks such as policy support stability risk, regulatory and approval risk, and market system change risk; economic-related risks such as investment cost control risk, financing and interest rate risk, and return uncertainty risk; market-related risks such as electricity price and demand fluctuation risk, competition and substitution risk, and market access and transaction risk; technology-related risks such as technology maturity risk, system integration and compatibility risk, and equipment reliability and performance risk; natural-related risks such as geological and gas storage safety risk, and climate and environmental adaptability risk; and operational risks such as safety production risk, operation and maintenance management level risk, control and dispatch risk, and supply chain and spare parts guarantee risk.
