Water environment treatment project combination multi-period selection method oriented to staged strategy agreement

Abstract

The invention discloses a water environment treatment project combination multi-period selection method oriented to staged strategy agreement, and particularly relates to the field of water environment treatment and project combination management.The method comprises the following steps that a strategic decomposition and priority evaluation method GC-W / VI KOR method is put forward according to the cross-period characteristic of a water environment treatment enterprise strategic target, and a multi-period selection result is obtained; decomposing a long-term strategic target of the organization, and obtaining an execution priority of each staged strategic target; designing a project flexible execution strategy and a resource flexible configuration strategy, and constructing a project combination selection model matched with a staging strategy under a resource limited condition; and proposing a simulation runoff mimicry algorithm, and solving the constructed project combination multi-period selection model.

Description

Technical Field

[0001] The present invention relates to the field of water environment management and project portfolio management, and in particular to a multi-phase selection method for water environment management project portfolios in accordance with phased strategic fit. Background Art

[0002] Water environment governance projects are of unique importance due to their dual strategic value of ecology and economy. The ecological value of the project is reflected in its role in water resource protection and pollution prevention, while the economic value is reflected in the return on investment of project implementation. However, the relationship between the two is complex, and the single pursuit of any one value may affect the long-term sustainable development of the enterprise. Therefore, when enterprises are conducting water environment governance, they need to scientifically select and combine multiple projects under limited resources to achieve a balance between ecological benefits and economic benefits, thereby ensuring the long-term development of the enterprise. From a strategic perspective, water environment governance involves multiple dimensions, and a single project is usually difficult to fully meet the needs of the enterprise. Therefore, the key challenge facing enterprise managers is to develop a project portfolio strategy that not only meets the overall strategic goals of water environment governance, but also brings stable economic returns to the enterprise and achieves sustainable development.

[0003] Water environment governance is a long-term and multi-stage governance process. Decomposing the long-term strategic goals of water environment governance into a series of executable phased strategic tasks is a necessary path for the successful implementation of water environment governance. Therefore, during the implementation stage of the water environment governance project, how the management can formulate an operational multi-phase selection plan for the project portfolio under the constraints of resources, organizational capabilities, etc., in order to achieve the phased strategic tasks at different stages, is a key issue to be discussed.

[0004] The research on project portfolio selection model for multi-stage strategic fit is still in its infancy. Most of the current research on project portfolio selection based on strategic fit only formulates project portfolio selection plans based on the long-term strategic goals of the organization. However, long-term strategic goals reflect the long-term vision of organizational development and are difficult to achieve in the short term. Especially for water environment governance projects, their investment payback period is far longer than that of general commercial investment projects, mostly 5 to 25 years. Therefore, it is inevitable to decompose long-term strategic goals into a series of executable phased strategic goals. Therefore, how to make multi-stage selections of project portfolios accordingly to complete the strategic tasks of each stage is the core of water environment governance project portfolio selection.

[0005] Currently, many studies on strategic project portfolio selection mainly focus on the initial planning stage, while relatively few studies are conducted on project portfolio selection during project implementation. Summary of the invention

[0006] To this end, the present invention provides a multi-phase selection method for water environment governance project portfolios that are compatible with phased strategies. The method combines the fuzziness and complexity of the project implementation process, proposes a strategic goal decomposition method based on the cloud fuzzy method, and transforms long-term strategic goals into orderly and executable phased strategic goals, thereby guiding the phased implementation process of the project portfolio. In addition, the present invention also introduces a flexible mechanism to further flexibly optimize the project portfolio implementation plan.

[0007] Water environment governance is a long-term and phased governance process, which means that the leading enterprise in water environment governance needs to achieve not only the overall strategic goal, but also the phased governance goals. Therefore, in the project implementation phase, it is necessary to use the multi-phase selection method of water environment governance project portfolio to seek a project portfolio implementation plan that can meet the strategic goals of different stages.

[0008] The present invention can help managers transform the long-term water environment governance process into a series of executable phased strategic tasks, and formulate more detailed and more operational project portfolio implementation plans for managers based on these phased tasks to solve the problems raised in the background technology.

[0009] The present invention provides the following technical solution: a multi-phase selection method for a water environment management project portfolio that matches the phased strategy, comprising the following steps:

[0010] S1: In view of the cross-cycle characteristics of the strategic goals of water environment management enterprises, the GC-W / VIKOR method of strategic decomposition and priority assessment is proposed to decompose the long-term strategic goals of the organization and obtain the execution priority of each phase of strategic goals;

[0011] S2: Design flexible project execution strategy and flexible resource allocation strategy, and build a project portfolio selection model that matches the phased strategy under resource constraints;

[0012] S3: A simulated runoff mimicry algorithm is proposed to solve the constructed project portfolio multi-period selection model.

[0013] Preferably, the GC-W / VIKOR method for decomposing the organization's long-term strategic goals and obtaining the priority of each phased strategic goal in S1 is specifically:

[0014] Long-term strategic goals can be broken down into executable phased strategic goals according to unit cycles T1 to T3. Enterprises select projects for combined implementation based on the priorities of phased strategic goals to meet current goals.

[0015] Taking T1 as an example, under the resource constraints, the enterprise selects and combines implementation projects to meet the current strategic goals x1, y1 and z1, and the priority of goal matching satisfies x1>y1>z1. This process is the multi-stage selection of project portfolios for phased strategic matching.

[0016] S11: Construct Gaussian cloud weighting method. The specific process is as follows:

[0017] S111: Establish priority criteria and set Gaussian cloud scale, specifically:

[0018] Through interviews with the operators of the leading enterprises, the priority assessment criteria are constructed, which are as follows: 1) D1: assessment priority of the central government department (National Development and Reform Commission); 2) D2: assessment priority of local competent departments; 3) D3: priority of social public demand; 4) D4: priority of enterprise development; 5) D5: priority of resource allocation; 6) D6: priority of economic benefits. Then, Gaussian cloud parameters (Ex, En, He) are defined to obtain cloud scale {C1~C9}, and cloud scale is used to assess the importance weights of different criteria;

[0019] Among them, D i is the priority evaluation standard, Ex represents the expected value of the Gaussian cloud distribution in the domain space, En is the entropy, He is the hyperentropy, and C i is the cloud scale;

[0020] S112: Construct a pairwise comparison matrix, specifically:

[0021] The cloud scale of {C1~C9} is used to construct the pairwise comparison matrix C for the criterion layer indicators for evaluating the priority of strategic goals. g ; Assuming the number of indicators in the criterion layer is k, then C g It is expressed as:

[0022]

[0023] In the formula, C g is the pairwise comparison matrix, k is the number of criteria layer indicators, and the comparison matrix C g The element C in ij represents the cloud scale importance of criterion i relative to criterion j, expressed as:

[0024]

[0025] S113: Further calculate the Gaussian cloud weights of different criterion layers, expressed as:

[0026]

[0027] Then, the square root method was used to calculate the cloud parameters in the Gaussian cloud weights

[0028] S114: Perform consistency check. If the matrix passes the consistency check, the Gaussian cloud weight set W of the criterion layer k is considered valid, then W k is represented as:

[0029]

[0030] S12: Gaussian cloud VIKOR model construction, the specific process is as follows:

[0031] S121: Construct a cloud-weighted strategic goal performance matrix, specifically:

[0032] Construct the strategic goal performance matrix Z without cloud weight v , which is expressed as follows:

[0033]

[0034] According to the cloud weight set of the criterion layer obtained in S114, the strategic goal performance matrix Z v Perform cloud weighted calculation, expressed as:

[0035]

[0036] According to the multiplication rule of cloud, calculate the cloud-weighted strategic goal performance matrix The performance value in the cloud multiplication rule is:

[0037]

[0038] Among them, A and B simply indicate that there are two different indicators;

[0039] Cloud-weighted strategic goal performance matrix It is expressed as:

[0040]

[0041] Cloud Weighted Strategic Objectives Performance Matrix Elements in satisfy:

[0042]

[0043] S122: Determine the cloud positive ideal value and the cloud negative ideal value, specifically:

[0044] Determine the cloud positive ideal value for each criterion layer Negative ideal value with cloud The calculation is as follows:

[0045]

[0046] In the formula, It represents the cloud-weighted performance value of the i-th (i∈m) phased strategic goal under the j-th (j∈k) criterion that represents the priority;

[0047] S123: Based on the calculated cloud positive ideal value and cloud negative ideal value, the cloud disadvantage degree S value and R value are further solved. The calculation method is as follows:

[0048]

[0049] In the formula, the R value represents the disadvantage of the strategic goal in the most important criterion, that is, the criterion with the largest cloud weight value. The lower the R value of the strategic goal, the higher the execution priority of the goal in the most important criterion. The S value represents the average disadvantage of the strategic goal in all criteria. The lower the S value, the higher the average execution priority of the goal in all criteria.

[0050] S124: Calculate the cloud Q value, specifically:

[0051] First, calculate the maximum and minimum cloud S and cloud R values ​​of all phased strategic goals, as follows:

[0052]

[0053] Furthermore, the comprehensive indicator cloud Q value of each phased strategic goal is calculated and expressed as:

[0054]

[0055] In the formula, w Q Represents the adjustment factor of cloud S value and cloud R value, set to w Q =0.5, which means that the cloud S value representing the local disadvantage and the cloud R value representing the global disadvantage are equally important;

[0056] S125: Cloud Q value defuzzification and calculation of execution priority, specifically:

[0057] The calculated cloud Q value is deblurred, and the calculation method is as follows:

[0058]

[0059] Where N c is the total number of cloud droplets in the Gaussian cloud, x n and n is the horizontal and vertical coordinate value of the corresponding cloud droplet, combined with the deblurred The execution priorities of phased strategic goals are sorted by value. The target with the smallest value has the highest execution priority.

[0060] Preferably, the construction process of the project portfolio selection model that matches the phased strategy in step S2 is:

[0061] S21: For the generality of model construction and analysis, the following assumptions are made:

[0062] H1: The unit cycle is a standard time unit, including years, quarters, months, weeks or their integer multiples. The long-term strategic goals are implemented in phases according to T unit cycles, forming T+1 time nodes. At the initial point, the enterprise only launches new projects without terminating any projects. At the end point, the project portfolio enters the termination state and no new projects are launched at this time.

[0063] H2: Each unit period is continuous, where t = 1 indicates the beginning of the first unit period, t = 2 indicates both the beginning of the second unit period and the end of the first strategic period, and so on;

[0064] H3: If the projects currently being implemented can meet the needs of the strategic goals, no new projects will be selected for implementation. Projects that are not selected for implementation will enter the candidate project pool and their implementation possibility will be considered again after entering the next unit cycle. All projects in the project portfolio must be completed within the entire strategic cycle.

[0065] H4: In the multi-period selection process of project portfolio, enterprise resources are allocated according to unit periods. To fit the reality, it is assumed that the amount of resources allocated in each unit period is flexible and variable, including leasing and purchasing, and the remaining resources that are not fully utilized in each unit period are automatically accumulated to the next strategic period.

[0066] S22: Model constraint construction, specifically:

[0067] S221: Regarding the execution constraints of the project's flexible execution strategy, specifically:

[0068] In the process of implementing water environment governance projects, considering the limitation of funds and resources, the project is postponed in a specific period of time. Through phased implementation, it can release some of the occupied resources. The enterprise can temporarily allocate these resources to other projects that are more critical to the implementation of the phased strategy, thereby effectively optimizing resource utilization and promoting the smooth achievement of strategic goals. Specifically, the project is completed in a certain proportion rather than in full in a specific time period. The range of project execution ratio is [α it ,β it ], then the completion ratio of the project in the unit strategic cycle satisfies: pr t,j ∈0∪[α it ,β it ], the multi-phase flexible execution constraints of the project are:

[0069] α t,j r t,j ≤pr t,j ≤β t,j r t,j ;

[0070] Among them, j represents a certain project;

[0071] S222: In the multi-period selection process of the water environment governance project portfolio, when the implementation requirements of the selected project portfolio exceed the limitations of the enterprise's own resources, the lead enterprise adopts a flexible resource allocation strategy, that is, the lead enterprise will obtain additional required resources through market-oriented means such as leasing or purchasing. Flexible resources are further divided into renewable flexible resources and non-renewable flexible resources. Multi-period flexible resource constraints are as follows:

[0072] ① Flexible non-renewable resource constraints:

[0073]

[0074] In the formula, Indicates that the enterprise has its own non-renewable resources. It refers to the flexible non-renewable resources that enterprises obtain through procurement and buyout;

[0075] ② Flexible renewable resource constraints:

[0076]

[0077] In the formula, It means that the enterprise has its own renewable resources. It refers to the flexible and renewable resources that enterprises obtain through deployment and leasing;

[0078] S223: Logical constraints on project execution:

[0079] Project execution needs to meet certain logical relationships. For example, the sewage treatment plant project can only be executed after the completion of related water supply system projects. First, define the logical relationship set E between projects. loc , if the context relationship between two items (i,j)∈E loc , it means that project j can only start after project i is completed. The logical constraints of project execution are expressed as:

[0080]

[0081] S224: Project completion constraints:

[0082] ∑ t∈T ∑ j∈J r t,j =x j ;

[0083] The project completion constraint states that if a project is selected, it must be completed within the entire investment period;

[0084] S225: Project executable quantity constraints

[0085] Due to the limitations of resources, funds, and company capabilities, the number of projects that can be selected and executed within the unit cycle and the entire strategic cycle is limited. Therefore, it is necessary to set clear limits on the number of selectable projects. The constraint on the number of projects within a unit cycle is expressed as:

[0086]

[0087] Where N is the maximum number of projects that can be carried out simultaneously during the entire strategic cycle;

[0088] S226: Multi-period water environment governance effect opportunity constraints:

[0089]

[0090] In the formula This is the water environment management standard value required by the government;

[0091] The water environment governance project portfolio implemented in each unit cycle must achieve the governance objectives set by the current public sector. The multi-period water environment governance effect constraint ensures that the implementation of the project portfolio can not only achieve the overall governance objectives, but also achieve the objectives of each stage. In actual scenarios, enterprises usually leave a certain margin considering the fluctuation of governance effects during the implementation of water environment comprehensive projects, that is, the governance effect of the projects actually implemented will be higher than the minimum limit required by the government.

[0092] S227: 0-1 constraints on decision variables

[0093] The 0-1 constraints corresponding to the decision variables are shown below:

[0094]

[0095] S23: Model objective function construction, specifically:

[0096] The phased strategic fit objective function and the profit objective function of the project portfolio multi-period selection model are shown below. In order to take these two objectives into account, the compromise satisfaction strategy is introduced to construct the objective function of the project portfolio multi-period selection model. The phased strategic fit objective and the profit objective are expressed as follows:

[0097]

[0098]

[0099] In the formula, and ρ m They represent the usage cost of unit flexible non-renewable resources and unit flexible renewable resources, DC j,t Q represents the additional cost incurred by flexible execution (delayed execution) of project j in period t, irepresents the Q value of the i-th phase strategic target calculated by the GC-W / VIKOR method, is the priority weight of the project, FR t,i Represents the phased project portfolio selection scheme r t,j The achievement rate of the i-th phased strategic goal in period t;

[0100] Then, the compromise satisfaction aggregation strategy is used to combine the phased strategy fit objective and the profit objective function into a single objective function. The specific process is as follows:

[0101] S231: Calculate the optimal value and the worst value of each objective respectively, expressed as: and The optimal and worst values ​​of the target value are regarded as the left and right boundaries of the satisfactory solution interval. For each target, its closeness to the optimal value is μ i It is expressed as:

[0102]

[0103] S232: The improved Zimmerman method is used to aggregate the closeness of the optimal values ​​of the two objectives. The model using the compromise satisfaction aggregation strategy is expressed as follows:

[0104]

[0105]

[0106] Where: W is the objective function value corresponding to the feasible solution that satisfies the model constraints, λ i (i∈{c,s}) is the psychological attention index, λ i The size of the reflection manager's attention mentality, i The closer it is to 1, the less attention is paid to this target value, and vice versa. SAT is the compromise satisfaction value of weighing the two goals.

[0107] The multi-period model of water environment governance project portfolio for phased strategic fit is expressed as follows:

[0108]

[0109]

[0110] Preferably, the solution process of the simulated runoff mimicry algorithm in step S3 is:

[0111] S31: The runoff algorithm first randomly generates a number of water flows to form a population. The calculation method of the initial position of the runoff is:

[0112] L X(i) =x L+r d ×(x H -x L );

[0113] Where, L X(i) represents the location of the runoff at the i-th location, x H and x L represents the upper and lower limits of the initial runoff position, r d is a random value uniformly distributed between 0 and 1. Then, it is assumed that there are J neighbors near each runoff location, and the location of the neighborhood is expressed as:

[0114] N X(j) =L X(j) +r dn ×Δ;

[0115] Where N X(j) represents the position of the jth neighbor (j∈J), r dn is a random value that obeys the normal distribution, Δ is the preset search space parameter. The larger the Δ value, the larger the search space of the runoff algorithm and the stronger the global optimization ability. The smaller the Δ value, the higher the accuracy of the local optimization. Therefore, determining the appropriate Δ value is the key to ensuring the performance of the algorithm. Otherwise, only pursuing global optimization will make it difficult for the algorithm to accurately locate the optimal value, while only pursuing local optimization will easily cause the algorithm to fall into the local optimal solution. When using the linear random strategy, the calculation method of the Δ value in the runoff algorithm is as follows:

[0116]

[0117] In the formula, By L X(i) The random position generated, the first term represents L X(i) Gradually move to a new random position The second term represents L X(i) Gradually move to the optimal position B through iteration X , so that B X and L X(i) The Euclidean distance of ||B X -L X(i) || gradually approaches 0, which makes the algorithm search space gradually shrink and tend to be stable. rd is a random weight strategy, calculated as:

[0118]

[0119] In the formula, is a uniformly distributed random vector, and W is a random weight strategy, which further introduces randomness into the runoff algorithm, thereby preventing the algorithm from falling into the local optimum;

[0120] In the runoff algorithm, the runoff velocity V is positively correlated with the slope of its location, so the flow velocity V is expressed as:

[0121] V=r dn ×g0(i,j,d);

[0122] Where g0 represents the slope of the runoff location, which is determined by:

[0123]

[0124] In the formula, F(·) represents the fitness value (altitude) of the location where the runoff is calculated, and the new location of the runoff is The update formula is expressed as:

[0125]

[0126] Then by iterating the runoff position Directing runoff to lower elevations Thus, the final outlet is found, which means that the runoff algorithm finds the global optimal solution;

[0127] S32: Use the Tent chaos strategy to generate the initial population, specifically:

[0128] The present invention uses Tent mapping to generate chaotic sequences, so that the initial solutions are distributed as evenly as possible in the solution space, aiming to enhance the optimization ability of the algorithm by improving the uniformity of the initial solutions. The method for generating chaotic sequences based on Tent mapping is as follows:

[0129]

[0130] In the formula, N p is the population number, I t is the current iteration number, according to the chaotic sequence Generate a sequence of runoff locations within the search area The process is as follows:

[0131]

[0132] In the formula, and They are The maximum and minimum values ​​of the sequence;

[0133] S33: Update the neighborhood flow position using the sinusoidal strategy, specifically:

[0134] In the runoff algorithm, since the search space operator Δ is linearly transformed in a random direction, the optimization efficiency of the algorithm is limited. Based on this, the present invention adopts a sinusoidal strategy to improve the operator Δ and adjusts the location update method of the neighborhood flow, thereby better balancing the global search ability and local optimization ability of the algorithm on the basis of ensuring the search space. The improved runoff neighborhood location update method and search space operator Δ * The calculation formula is as follows:

[0135] N X(j) =L X(i) ×Δ * ;

[0136]

[0137] Where R1 and R2 are random angle numbers that determine the displacement distance and displacement direction in the next iteration, and σ1 and σ2 are based on the golden section coefficients. The improved runoff algorithm improves the optimization efficiency in the algorithm search space and ensures the convergence of the algorithm by introducing the sine strategy.

[0138] The present invention has the following advantages:

[0139] 1. In order to reduce or eliminate the adverse interference of changes in uncertain factors on the strategic priority assessment of leading enterprises in water environment management, the present invention designs the GC-W / VIKOR method based on the fuzzy decision-making principle to achieve the priority assessment of the phased strategic goals of leading enterprises in water environment management. Leading enterprises in water environment management face a long investment payback cycle, a wide range of project implementation, unpredictable market prospects and complex operating conditions. This complexity brings great uncertainty to the strategic decision-making of enterprises. Under the interference of uncertain factors, the strategic decision-making of enterprises may deviate. Therefore, leading enterprises in water environment management need to fully consider potential uncertainties when assessing strategic priorities.

[0140] 2. The present invention introduces the cloud model to characterize the uncertainty in strategic priority assessment, and it is necessary to measure the uncertainty. Randomness and fuzziness are the basic characteristics of uncertainty. Considering that the cloud model (CM) can effectively characterize randomness and fuzziness and the correlation between the two, it can reduce or eliminate the interference of uncertainty factors on strategic priority assessment.

[0141] Taking into account the possible conflicts among the evaluation criteria in the strategic priority assessment criterion layer and the ambiguity and randomness in the assessment process, the present invention improves the classic VIKOR algorithm through the Gaussian cloud model and proposes a Gaussian cloud-VIKOR sorting algorithm to assess the priority of each strategic goal in each phase. BRIEF DESCRIPTION OF THE DRAWINGS

[0142] Figure 1 A multi-stage strategic goal breakdown diagram provided for the present invention;

[0143] Figure 2 The long-term strategic goal provided by the present invention is decomposed into a process diagram of phased strategic goals;

[0144] Figure 3 A schematic diagram of the priority assessment of the phased strategic objectives provided by the present invention;

[0145] Figure 4 A schematic diagram of a multi-phase selection process for a project portfolio based on phased strategic goal alignment provided by the present invention;

[0146] Figure 5 A flow chart of the GC-W / VIKOR method provided by the present invention;

[0147] Figure 6 A process diagram of the phased strategic goal priority assessment based on GC-W / VIKOR provided by the present invention;

[0148] Figure 7 A diagram of the multi-stage selection process of a project portfolio provided by the present invention;

[0149] Figure 8 A schematic diagram of a multi-phase flexible strategy for a project portfolio provided by the present invention;

[0150] Fig. 9 Flow chart of the RO-CG algorithm provided by the present invention;

[0151] Fig.10 Cloud data features for evaluating language values ​​provided by the present invention;

[0152] Fig.11 A three-dimensional graph of the benchmark function provided by the present invention;

[0153] Fig.12 A solution process curve of the algorithm on the benchmark test function provided by the present invention;

[0154] Fig.13 The multi-phase selection scheme for the combined phase I CT project of water environment treatment provided by the present invention;

[0155] Fig.14 This is the water environment management effect of the water environment management CT Phase I project combination provided by the present invention. DETAILED DESCRIPTION

[0156] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0157] like Figure 1-5 As shown, the present invention provides a multi-phase selection method for a water environment management project portfolio that matches a phased strategy, comprising the following steps:

[0158] S1: In view of the cross-cycle characteristics of the strategic goals of water environment management enterprises, the GC-W / VIKOR method of strategic decomposition and priority assessment is proposed to decompose the long-term strategic goals of the organization and obtain the execution priority of each phase of strategic goals;

[0159] Long-term strategic goals can be broken down into executable phased strategic goals according to unit cycles T1 to T3. Enterprises select projects for combined implementation based on the priorities of phased strategic goals to meet current goals.

[0160] Taking T1 as an example, under the resource constraints, the enterprise selects and combines implementation projects to meet the current strategic goals x1, y1 and z1, and the priority of goal matching satisfies x1>y1>z1. This process is the multi-stage selection of project portfolios for phased strategic matching.

[0161] S11: Construct Gaussian cloud weighting method. The specific process is as follows:

[0162] S111: Establish priority criteria and set Gaussian cloud scale, specifically:

[0163] Through interviews with the operators of the leading enterprises, the priority assessment criteria are constructed, which are as follows: 1) D1: assessment priority of the central government department (National Development and Reform Commission); 2) D2: assessment priority of local competent departments; 3) D3: priority of social public demand; 4) D4: priority of enterprise development; 5) D5: priority of resource allocation; 6) D6: priority of economic benefits. Then, Gaussian cloud parameters (Ex, En, He) are defined to obtain cloud scale {C1~C9}, and cloud scale is used to assess the importance weights of different criteria;

[0164] Among them, D i is the priority evaluation standard, Ex represents the expected value of the Gaussian cloud distribution in the domain space, En is the entropy, He is the hyperentropy, and C i is the cloud scale;

[0165] S112: Construct a pairwise comparison matrix, specifically:

[0166] The cloud scale of {C1~C9} is used to construct the pairwise comparison matrix C for the criterion layer indicators for evaluating the priority of strategic goals. g ; Assuming the number of indicators in the criterion layer is k, then C g It is expressed as:

[0167]

[0168] In the formula, C g is the pairwise comparison matrix, k is the number of criteria layer indicators, and the comparison matrix C g The element C in ij represents the cloud scale importance of criterion i relative to criterion j, expressed as:

[0169]

[0170] S113: Further calculate the Gaussian cloud weights of different criterion layers, expressed as:

[0171]

[0172] Then, the square root method was used to calculate the cloud parameters in the Gaussian cloud weights

[0173] S114: Perform consistency check. If the matrix passes the consistency check, the Gaussian cloud weight set W of the criterion layer k is considered valid, then W k is represented as:

[0174]

[0175] S12: Gaussian cloud VIKOR model construction, the specific process is as follows:

[0176] S121: Construct a cloud-weighted strategic goal performance matrix, specifically:

[0177] Construct the strategic goal performance matrix Z without cloud weight v , which is expressed as follows:

[0178]

[0179] According to the cloud weight set of the criterion layer obtained in S114, the strategic goal performance matrix Z v Perform cloud weighted calculation, expressed as:

[0180]

[0181] According to the multiplication rule of cloud, calculate the cloud-weighted strategic goal performance matrix The performance value in the cloud multiplication rule is:

[0182]

[0183] Among them, A and B simply indicate that there are two different indicators;

[0184] Cloud-weighted strategic goal performance matrix It is expressed as:

[0185]

[0186] Cloud Weighted Strategic Objectives Performance Matrix Elements in satisfy:

[0187]

[0188] S122: Determine the cloud positive ideal value and the cloud negative ideal value, specifically:

[0189] Determine the cloud positive ideal value for each criterion layer Negative ideal value with cloud The calculation is as follows:

[0190]

[0191] In the formula, It represents the cloud-weighted performance value of the i-th (i∈m) phased strategic goal under the j-th (j∈k) criterion that represents the priority;

[0192] S123: Based on the calculated cloud positive ideal value and cloud negative ideal value, the cloud disadvantage degree S value and R value are further solved. The calculation method is as follows:

[0193]

[0194]

[0195] In the formula, the R value represents the disadvantage of the strategic goal in the most important criterion, that is, the criterion with the largest cloud weight value. The lower the R value of the strategic goal, the higher the execution priority of the goal in the most important criterion. The S value represents the average disadvantage of the strategic goal in all criteria. The lower the S value, the higher the average execution priority of the goal in all criteria.

[0196] S124: Calculate the cloud Q value, specifically:

[0197] First, calculate the maximum and minimum cloud S and cloud R values ​​of all phased strategic goals, as follows:

[0198]

[0199] Furthermore, the comprehensive indicator cloud Q value of each phased strategic goal is calculated and expressed as:

[0200]

[0201] In the formula, w Q Represents the adjustment factor of cloud S value and cloud R value, set to w Q =0.5, which means that the cloud S value representing the local disadvantage and the cloud R value representing the global disadvantage are equally important;

[0202] S125: Cloud Q value defuzzification and calculation of execution priority, specifically:

[0203] The calculated cloud Q value is deblurred, and the calculation method is as follows:

[0204]

[0205] Where N c is the total number of cloud droplets in the Gaussian cloud, x n and n is the horizontal and vertical coordinate value of the corresponding cloud droplet, combined with the deblurred The execution priorities of phased strategic goals are sorted by value. The target with the smallest value has the highest execution priority.

[0206] S2: Design flexible project execution strategies and flexible resource allocation strategies, and build a project portfolio selection model that matches the phased strategy under resource constraints; Figure 6 , Figure 7 and Figure 8 As shown in Figure 1, the process of building a project portfolio selection model is as follows:

[0207] S21: For the generality of model construction and analysis, the following assumptions are made:

[0208] H1: The unit cycle is a standard time unit, including years, quarters, months, weeks or their integer multiples. The long-term strategic goals are implemented in phases according to T unit cycles, forming T+1 time nodes. At the initial point, the enterprise only launches new projects without terminating any projects. At the end point, the project portfolio enters the termination state and no new projects are launched at this time.

[0209] H2: Each unit period is continuous, where t = 1 indicates the beginning of the first unit period, t = 2 indicates both the beginning of the second unit period and the end of the first strategic period, and so on;

[0210] H3: If the projects currently being implemented can meet the needs of the strategic goals, no new projects will be selected for implementation. Projects that are not selected for implementation will enter the candidate project pool and their implementation possibility will be considered again after entering the next unit cycle. All projects in the project portfolio must be completed within the entire strategic cycle.

[0211] H4: In the multi-period selection process of project portfolio, enterprise resources are allocated according to unit periods. To fit the reality, it is assumed that the amount of resources allocated in each unit period is flexible and variable, including leasing and purchasing, and the remaining resources that are not fully utilized in each unit period are automatically accumulated to the next strategic period.

[0212] S22: Model constraint construction, specifically:

[0213] S221: Regarding the execution constraints of the project's flexible execution strategy, specifically:

[0214] In the process of implementing water environment governance projects, considering the limitation of funds and resources, the project is postponed in a specific period of time. Through phased implementation, it can release some of the occupied resources. The enterprise can temporarily allocate these resources to other projects that are more critical to the implementation of the phased strategy, thereby effectively optimizing resource utilization and promoting the smooth achievement of strategic goals. Specifically, the project is completed in a certain proportion rather than all in a specific time period. The range of project execution ratio is [α it ,β it ], then the completion ratio of the project in the unit strategic cycle satisfies: pr t,j ∈0∪[α it ,β it ], the multi-phase flexible execution constraints of the project are:

[0215] α t,j r t,j ≤pr t,j ≤β t,j r t,j ;

[0216] Among them, j represents a certain project;

[0217] S222: In the multi-period selection process of the water environment governance project portfolio, when the implementation requirements of the selected project portfolio exceed the limitations of the enterprise's own resources, the lead enterprise adopts a flexible resource allocation strategy, that is, the lead enterprise will obtain additional required resources through market-oriented means such as leasing or purchasing. Flexible resources are further divided into renewable flexible resources and non-renewable flexible resources. Multi-period flexible resource constraints are as follows:

[0218] ① Flexible non-renewable resource constraints:

[0219]

[0220] In the formula, Indicates that the enterprise has its own non-renewable resources. It refers to the flexible non-renewable resources that enterprises obtain through procurement and buyout;

[0221] ② Flexible renewable resource constraints:

[0222]

[0223] In the formula, It means that the enterprise has its own renewable resources. It refers to the flexible and renewable resources that enterprises obtain through deployment and leasing;

[0224] S223: Logical constraints on project execution:

[0225] Project execution needs to meet certain logical relationships. For example, the sewage treatment plant project can only be executed after the completion of related water supply system projects. First, define the logical relationship set E between projects. loc , if the context relationship between two items (i,j)∈E loc , it means that project j can only start after project i is completed. The logical constraints of project execution are expressed as:

[0226]

[0227] S224: Project completion constraints:

[0228] ∑ t∈T ∑ j∈J r t,j =x j ;

[0229] The project completion constraint states that if a project is selected, it must be completed within the entire investment period;

[0230] S225: Project executable quantity constraints

[0231] Due to the limitations of resources, funds, and company capabilities, the number of projects that can be selected and executed within the unit cycle and the entire strategic cycle is limited. Therefore, it is necessary to set clear limits on the number of selectable projects. The constraint on the number of projects within a unit cycle is expressed as:

[0232]

[0233] Where N is the maximum number of projects that can be carried out simultaneously during the entire strategic cycle;

[0234] S226: Multi-period water environment governance effect opportunity constraints:

[0235]

[0236] In the formula This is the water environment management standard value required by the government;

[0237] The water environment governance project portfolio implemented in each unit cycle must achieve the governance objectives set by the current public sector. The multi-period water environment governance effect constraint ensures that the implementation of the project portfolio can not only achieve the overall governance objectives, but also achieve the objectives of each stage. In actual scenarios, enterprises usually leave a certain margin considering the fluctuation of governance effects during the implementation of water environment comprehensive projects, that is, the governance effect of the projects actually implemented will be higher than the minimum limit required by the government.

[0238] S227: 0-1 constraints on decision variables

[0239] The 0-1 constraints corresponding to the decision variables are shown below:

[0240]

[0241] S23: Model objective function construction, specifically:

[0242] The phased strategic fit objective function and the profit objective function of the project portfolio multi-period selection model are shown below. In order to take these two objectives into account, the compromise satisfaction strategy is introduced to construct the objective function of the project portfolio multi-period selection model. The phased strategic fit objective and the profit objective are expressed as follows:

[0243]

[0244]

[0245] In the formula, and ρ m They represent the usage cost of unit flexible non-renewable resources and unit flexible renewable resources, DC j,t Q represents the additional cost incurred by flexible execution (delayed execution) of project j in period t, i represents the Q value of the i-th phase strategic target calculated by the GC-W / VIKOR method, is the priority weight of the project, FR t,i Represents the phased project portfolio selection scheme r t,j The achievement rate of the i-th phased strategic goal in period t;

[0246] Then, the compromise satisfaction aggregation strategy is used to combine the phased strategy fit objective and the profit objective function into a single objective function. The specific process is as follows:

[0247] S231: Calculate the optimal value and the worst value of each objective respectively, expressed as: and The optimal and worst values ​​of the target value are regarded as the left and right boundaries of the satisfactory solution interval. For each target, its closeness to the optimal value is μ i It is expressed as:

[0248]

[0249] S232: The improved Zimmerman method is used to aggregate the closeness of the optimal values ​​of the two objectives. The model using the compromise satisfaction aggregation strategy is expressed as follows:

[0250]

[0251]

[0252] Where: W is the objective function value corresponding to the feasible solution that satisfies the model constraints, λ i (i∈{c,s}) is the psychological attention index, λ i The size of the reflection manager's attention mentality, i The closer it is to 1, the less attention is paid to this target value, and vice versa. SAT is the compromise satisfaction value of weighing the two goals.

[0253] The multi-period model of water environment governance project portfolio for phased strategic fit is expressed as follows:

[0254]

[0255]

[0256] S3: A simulated runoff mimicry algorithm is proposed to solve the constructed project portfolio multi-period selection model, such as Fig. 9 As shown, the solution process is:

[0257] S31: The runoff algorithm first randomly generates a number of water flows to form a population. The calculation method of the initial position of the runoff is:

[0258] L X(i) =x L +r d ×(x H -x L );

[0259] Where, L X(i) represents the location of the runoff at the i-th location, x H and x L represents the upper and lower limits of the initial runoff position, r d is a random value uniformly distributed between 0 and 1. Then, it is assumed that there are J neighbors near each runoff location, and the location of the neighborhood is expressed as:

[0260] N X(j) =L X(i) +r dn ×Δ;

[0261] Where, LX(j) represents the position of the jth neighbor (j∈J), r dn is a random value that obeys the normal distribution, Δ is the preset search space parameter. The larger the Δ value, the larger the search space of the runoff algorithm and the stronger the global optimization ability. The smaller the Δ value, the higher the accuracy of the local optimization. Therefore, determining the appropriate Δ value is the key to ensuring the performance of the algorithm. Otherwise, only pursuing global optimization will make it difficult for the algorithm to accurately locate the optimal value, while only pursuing local optimization will easily cause the algorithm to fall into the local optimal solution. When using the linear random strategy, the calculation method of the Δ value in the runoff algorithm is as follows:

[0262]

[0263] In the formula, By L X(i) The random position generated, the first term represents L X(i) Gradually move to a new random position The second term represents L X(i) Gradually move to the optimal position B through iteration X , so that B X and L X(i) The Euclidean distance of ||B X -L X(i) || gradually approaches 0, which makes the algorithm search space gradually shrink and tend to be stable. rd is a random weight strategy, calculated as:

[0264]

[0265] In the formula, is a uniformly distributed random vector, and W is a random weight strategy, which further introduces randomness into the runoff algorithm, thereby preventing the algorithm from falling into a local optimum;

[0266] In the runoff algorithm, the runoff velocity V is positively correlated with the slope of its location, so the velocity V is expressed as:

[0267] V=r dn ×g0(i,j,d);

[0268] Where g0 represents the slope of the runoff location, which is determined by:

[0269]

[0270] In the formula, F(·) represents the fitness value (altitude) of the location where the runoff is calculated, and the new location of the runoff is The update formula is expressed as:

[0271]

[0272] Then by iterating the runoff position Directing runoff to lower elevations Thus, the final outlet is found, which means that the runoff algorithm finds the global optimal solution;

[0273] S32: Use the Tent chaos strategy to generate the initial population, specifically:

[0274] The present invention uses Tent mapping to generate chaotic sequences, so that the initial solutions are distributed as evenly as possible in the solution space, aiming to enhance the optimization ability of the algorithm by improving the uniformity of the initial solutions. The method for generating chaotic sequences based on Tent mapping is as follows:

[0275]

[0276] In the formula, N p is the population number, I t is the current iteration number, according to the chaotic sequence Generate a sequence of runoff locations within the search area The process is as follows:

[0277]

[0278] In the formula, and They are The maximum and minimum values ​​of the sequence;

[0279] S33: Update the neighborhood flow position using the sinusoidal strategy, specifically:

[0280] In the runoff algorithm, since the search space operator Δ is linearly transformed in a random direction, the optimization efficiency of the algorithm is limited. Based on this, the present invention adopts a sinusoidal strategy to improve the operator Δ and adjusts the location update method of the neighborhood flow, thereby better balancing the global search ability and local optimization ability of the algorithm on the basis of ensuring the search space. The improved runoff neighborhood location update method and search space operator Δ * The calculation formula is as follows:

[0281] N X(j) =L X(i )×Δ * ;

[0282]

[0283] Where R1 and R2 are random angle numbers that determine the displacement distance and displacement direction in the next iteration, and σ1 and σ2 are based on the golden section coefficients. The improved runoff algorithm improves the optimization efficiency in the algorithm search space and ensures the convergence of the algorithm by introducing the sine strategy.

[0284] To further illustrate the advancedness of the method of the present invention, examples are introduced to prove it:

[0285] 1. Background of the Embodiments

[0286] Taking the CT Phase I project portfolio of water environment governance that CJ Environmental Protection Group plans to invest in as an example, the total investment is expected to be about 455 million yuan. Through preliminary investigations and evaluations of the market, resources and government policies, 21 projects have been included in the candidate project pool. Some of these projects come from government review projects, and the other part comes from corporate inspection projects. Projects from both sources are included in the candidate project pool. CJ Environmental Protection Group needs to evaluate the candidate projects based on their scale, implementation capabilities, etc. The main environmental governance indicators obtained through the investigation are shown in Table 3, specifically: ① Sewage treatment volume (10,000 t / d); ② Flood retention zone landscaping project area (m 2 );③ Riverbank and lakeside ecological wetland restoration area (m 2 );④ Storage capacity (m 3 ) ; ⑤ Length of newly built and repaired sewage interception pipelines (km). The first phase of the CT project portfolio for water environment governance involves sewage treatment plants, urban sewage pipe network projects, inner lake source control and sewage interception projects, ecological wetland restoration projects, tailwater purification projects, and live water quality improvement projects. Due to limited funds and organizational capacity, only some projects that best match the strategic goals can be selected from the pool of alternative projects for combined implementation. After market research and investigation, the information of the alternative projects to be implemented (the alternative project numbers are: P1~P21).

[0287] Based on the above information, CJ Environmental Group plans to select several projects from the 21 candidate projects to form a water environment governance project portfolio for implementation, among which Project P1 is a government review project that the company must implement, and the remaining projects are company inspection projects that can be selectively implemented. Detailed project information and various parameters are shown in Tables 1 to 3.

[0288] Table 1 Information on the candidate projects for the first phase of CT water environment management of CJ Environmental Protection Group

[0289]

[0290]

[0291]

[0292] Table 2 Parameters of alternative projects for the first phase of the CT water environment treatment project

[0293]

[0294] Table 3 Phased demand for water environment governance in the public sector

[0295]

[0296] 2. Analysis of the process of prioritizing strategic goals in different phases

[0297] Based on project experience, resources required for goal execution, etc., the Gaussian cloud weighting method (GC-W) is used to evaluate the weights of the criterion layer in order to carry out the priority assessment of strategic goals in the future. According to the GC-W method proposed in Chapter 4, the pairwise comparison matrix and consistency test results are shown in Table 4.

[0298] Table 4. Consistency test results

[0299]

[0300] The results in Table 4 show that the above matrices all pass the consistency test. Therefore, the weights of the priority evaluation criteria layer can be calculated, as shown in Table 5.

[0301] Table 5 Priority assessment criteria layer weights

[0302]

[0303] After weighting the priority assessment criteria layer through the GC-W method and expert opinions, the Gaussian Cloud-VIKOR (GC-VIKOR) method is further combined to assess the execution priority of each phased strategic goal of each CJ Environmental Protection Group. The five-granularity language value and the corresponding cloud variable are used for the priority assessment process. The evaluation language value and the corresponding cloud variable are shown in Table 6. The corresponding Gaussian cloud data features are as follows Fig.10 shown.

[0304] Table 6 Evaluation language values ​​and corresponding cloud variables

[0305]

[0306] In the example of the CJ Environmental Protection Group's CT Phase I water environment management project portfolio, the results of using the GC-VIKOR method to assess the priorities of phased strategic objectives are shown in Table 7. A sensitivity analysis of the priority results was further conducted to verify its robustness.

[0307] Table 7 Priority assessment results of phased strategic objectives based on the GC-VIKOR method

[0308]

[0309]

[0310] The present invention adjusts w Q The sensitivity analysis of the results of the phased strategy priority was performed using the values ​​(0.1 to 0.9), as shown in Table 7. QThe value is set to 0.1-0.5, and the execution priority of the phased strategic goals r1-r4 is: r2>r3>r4>r1. That is, goal r2 should be allocated resources first and ensure its smooth implementation, while the importance of goal r1 is lower and the implementation order can be relatively later. Q The value of is increased from 0.1 to 0.5 to 0.6 to 0.9. At this time, the execution priority of target r3 exceeds that of target r2, and the overall priority order becomes: r3>r2>r4>r1, that is, target r2 should be executed first, and the priority order of other targets remains unchanged. It can be seen that w Q The setting of the value has a certain impact on the results of priority assessment, but the overall fluctuation of the results is not large, which shows the robustness of the GC-VIKOR method.

[0311] 3. Verification of solution algorithm performance

[0312] This paper uses 9 different benchmark functions to compare the performance differences between the RO-CG algorithm and the current mainstream optimization algorithms. The comparison algorithms are: Particle Swarm Optimization (PSO), Simulated Annealing (SA), Genetic Algorithm (GA), Antlion Optimization (ALO), Multiverse Optimization Algorithm (MVO), Sine Cosine Algorithm (SCA). Fig.11 A three-dimensional image representing a benchmark function. This embodiment uses both unimodal and multimodal benchmark functions to test the performance of the algorithm. In addition, only benchmark functions that show obvious performance differences are selected for testing (F1-F6, F8, F10, and F11), and functions that do not show obvious performance differences are not selected (F7, F9, F12, and F13).

[0313] Table 8 Algorithm test results on benchmark functions

[0314]

[0315]

[0316] In order to intuitively show the performance difference between the proposed RO-CG algorithm and other comparison algorithms, Fig.12The solution process curves of the RO-CG algorithm and other optimization algorithms are shown (the number of iterations is 20 times). Table 8 compares the result differences between the RO-CG algorithm and other optimization algorithms on the benchmark function (the number of iterations is 100 times). In order to obtain the true performance of the algorithm, each algorithm is independently run 20 times and the average value is recorded. As shown in Table 8, the RO-CG algorithm obtains the best results in 7 of the 9 benchmark test functions (F1-F4, F6, F10 and F11), and obtains suboptimal results in the remaining 2 benchmark test functions (F5 and F8), indicating the effectiveness of the designed RO-CG algorithm in the face of unimodal and multimodal optimization problems. Therefore, the present invention will use the RO-CG algorithm to solve the constructed project portfolio multi-period selection model.

[0317] 4. Analysis of multi-phase selection results of project portfolios that match the phased strategy

[0318] According to the execution priority order of the above-mentioned phased strategic goals, the proposed RO-CG algorithm is used to solve the multi-phase selection model of water environment governance project portfolio, and the optimal project portfolio multi-phase selection scheme that achieves phased strategic fit can be obtained. The scheme is shown in Fig.13 As shown in Figure 2, the compromise satisfaction of the optimal project portfolio multi-period selection scheme is 0.92. Fig.13 In the figure, the red dotted line represents the project execution cycle extended by adopting a flexible project execution strategy.

[0319] Depend on Fig.13 It can be seen that in order to achieve the goal of phased strategic goals, the optimal multi-phase selection plan for the project portfolio is: T1: {P1-P17-P18}; T2: {P4-P12-P15}; T3: {P11-P16-P21}; T4: {P2-P5-P10-P19}; T5: {P7-P9}. Project P1 (LF sewage comprehensive treatment project) and Project P18 (DY Lake sewage comprehensive treatment project) are given priority because they can obtain sufficient cash flow and promote the construction of subsequent projects. Project P7 (BL Lake Pump Station Project) and Project P9 (BL Lake Gate Project) have a lower degree of fit with the strategic goals and are difficult to generate stable cash flow, so their execution priority is lower, which shows that the obtained multi-phase selection plan for the project portfolio has a certain rationality. Subsequently, the results of the scheme are analyzed in detail. In Chapter 4, this embodiment uses a compromise satisfaction strategy to achieve the aggregation of the project portfolio revenue target and the phased strategy fit target. Therefore, it is necessary to analyze the compromise satisfaction strategy. First, define the project portfolio revenue reduction coefficient SCR, and its calculation formula is:

[0320]

[0321] In the formula, The lower bound of the project portfolio return obtained to maximize the phased strategy fit target. c is the lower bound of the current project portfolio return. For example, when the compromise satisfaction is 0.91, the expected change range of flexible investment is [0, 2.3%], indicating that if managers want to fully achieve all current phased strategic goals, they need to add 2.3% of additional flexible investment on the current basis. Therefore, the compromise satisfaction strategy adopted can guide managers to make effective flexible investments to achieve the current phased strategic goals to the greatest extent.

[0322] Further analysis of the water environment governance effects of the multi-phase selection scheme of the project portfolio, Fig.14 The figure shows the comparison between the water environment governance effect (blue curve) of the CT Phase I project portfolio under different time periods and the governance requirements (orange part) set by the government. Under certain constraints, the governance effect of the CT Phase I project portfolio will be close to the governance requirements of the public sector. Therefore, once the governance capacity of the project fluctuates, such as the sewage treatment equipment effect is not as good as expected, the ecological wetland restoration effect is not good, etc., the governance effect of the project portfolio will not be able to meet the requirements of the public sector, resulting in breach of contract, damage to corporate reputation, etc. The water environment governance opportunity constraint mechanism introduced in this article can significantly reduce the risk of substandard governance effects, and the higher the confidence level, the greater the margin between the project governance effect and the public sector governance requirements, that is, the smaller the risk of default. Fig.14 As shown, the multi-phase selection scheme of the project portfolio constructed by the present invention retains a reasonable margin with the public sector in each time period, thereby reducing the risk of substandard governance effects. It can be seen that the opportunity constraint mechanism helps to improve the water environment governance efficiency of the project portfolio and ensure the public service quality of the lead enterprise.

[0323] The present invention aims at the problems that water environment governance is characterized by long cycles and multiple stages, the strategic fit of project portfolio selection is low, and it is not conducive to the sustainable development of enterprises. First, facing the implementation stage of water environment governance projects, the basic connotations of strategic decomposition and strategic priority are explained, and the GC-W / VIKOR method is developed, aiming to formulate a strategic decomposition and strategic execution priority assessment plan for the leading enterprise of water environment governance. Secondly, combined with the proposed project flexible execution and resource flexible allocation strategies, a multi-stage selection model for water environment governance project portfolios with phased strategic fit is constructed. The model takes into account the multi-stage dynamic matching of the strategic goal decomposition process and the project portfolio selection process, and enhances the mobility and phased response capabilities of project portfolio management. Finally, the designed RO-CG algorithm is used to solve the model, thereby determining the optimal multi-stage selection plan for the project portfolio, thereby ensuring that the project portfolio can be smoothly implemented in stages and can maximize the realization of each stage strategic goal. At the same time, the high reuse and extensibility of the method are achieved, and only a small modification of some contents in the method is required to apply it to other complex environments.

[0324] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A multi-phase selection method for water environment management project portfolios that is compatible with phased strategies, characterized by: The following steps are involved: S1: In view of the cross-cycle characteristics of the strategic goals of water environment management enterprises, the GC-W / VIKOR method of strategic decomposition and priority assessment is proposed to decompose the long-term strategic goals of the organization and obtain the execution priority of each phase of strategic goals; S2: Design flexible project execution strategy and flexible resource allocation strategy, and build a project portfolio selection model that matches the phased strategy under resource constraints; S3: A simulated runoff mimicry algorithm is proposed to solve the constructed project portfolio multi-period selection model.

2. The multi-phase selection method for water environment management project portfolio with phased strategic fit according to claim 1 is characterized by: The GC-W / VIKOR method in S1 for decomposing the organization's long-term strategic goals and obtaining the priority of each phased strategic goal is specifically: Long-term strategic goals can be broken down into executable phased strategic goals according to unit cycles T1 to T3. Enterprises select projects for combined implementation based on the priorities of phased strategic goals to meet current goals. Taking T1 as an example, under the resource constraints, the enterprise selects and combines implementation projects to meet the current strategic goals x1, y1 and z1, and the priority of goal matching satisfies x1>y1>z1. This process is the multi-stage selection of project portfolios for phased strategic matching. S11: Construct Gaussian cloud weighting method. The specific process is as follows: S111: Establish priority criteria and set Gaussian cloud scale, specifically: Through interviews with the lead enterprise operators, we established the priority assessment criteria, which are as follows: 1) D1: Assessment priority of central government departments; 2) D2: Assessment priority of local competent authorities; 3) D3: Priority of public needs; 4) D4: Enterprise development priority; 5) D5: resource allocation priority; 6) D6: Priority of economic benefits; Then, Gaussian cloud parameters (Ex, En, He) are defined to obtain cloud scales {C1-C9}, and the cloud scales are used to assess the importance weights of different criteria; Among them, D i is the priority evaluation standard, Ex represents the expected value of the Gaussian cloud distribution in the domain space, En is the entropy, He is the hyperentropy, and C i is the cloud scale; S112: Construct a pairwise comparison matrix, specifically: The cloud scale of {C1~C9} is used to construct the pairwise comparison matrix C for the criterion layer indicators for evaluating the priority of strategic goals. g ; Assuming the number of indicators in the criterion layer is k, then C g It is expressed as: In the formula, C g is the pairwise comparison matrix, k is the number of criteria layer indicators, and the comparison matrix C g The element C in ij represents the cloud scale importance of criterion i relative to criterion j, expressed as: S113: Further calculate the Gaussian cloud weights of different criterion layers, expressed as: Then, the square root method was used to calculate the cloud parameters in the Gaussian cloud weights S114: Perform consistency check. If the matrix passes the consistency check, the Gaussian cloud weight set W of the criterion layer k is considered valid, then W k is represented as: S12: Gaussian cloud VIKOR model construction, the specific process is as follows: S121: Construct a cloud-weighted strategic goal performance matrix, specifically: Construct the strategic goal performance matrix Z without cloud weight v , which is expressed as follows: According to the cloud weight set of the criterion layer obtained in S114, the strategic goal performance matrix Z v Perform cloud weighted calculation, expressed as: According to the multiplication rule of cloud, calculate the cloud-weighted strategic goal performance matrix The performance value in the cloud multiplication rule is: Among them, A and B simply indicate that there are two different indicators; Cloud-weighted strategic goal performance matrix It is expressed as: Cloud Weighted Strategic Objectives Performance Matrix Elements in satisfy: S122: Determine the cloud positive ideal value and the cloud negative ideal value, specifically: Determine the cloud positive ideal value for each criterion layer Negative ideal value with cloud The calculation is as follows: In the formula, It represents the cloud-weighted performance value of the i-th (i∈m) phased strategic goal under the j-th (j∈k) criterion that represents the priority; S123: Based on the calculated cloud positive ideal value and cloud negative ideal value, the cloud disadvantage degree S value and R value are further solved. The calculation method is as follows: In the formula, the R value represents the disadvantage of the strategic goal in the most important criterion, that is, the criterion with the largest cloud weight value. The lower the R value of the strategic goal, the higher the execution priority of the goal in the most important criterion. The S value represents the average disadvantage of the strategic goal in all criteria. The lower the S value, the higher the average execution priority of the goal in all criteria. S124: Calculate the cloud Q value, specifically: First, calculate the maximum and minimum cloud S and cloud R values ​​of all phased strategic goals, as follows: Furthermore, the comprehensive indicator cloud Q value of each phased strategic goal is calculated and expressed as: In the formula, w Q Represents the adjustment factor of cloud S value and cloud R value, set to w Q =0.5, which means that the cloud S value representing the local disadvantage and the cloud R value representing the global disadvantage are equally important; S125: Cloud Q value defuzzification and calculation of execution priority, specifically: The calculated cloud Q value is deblurred, and the calculation method is as follows: Where N c is the total number of cloud droplets in the Gaussian cloud, x n and n is the horizontal and vertical coordinate value of the corresponding cloud droplet, combined with the deblurred The execution priorities of phased strategic goals are sorted by value. The target with the smallest value has the highest execution priority.

3. The multi-phase selection method for water environment management project portfolio with phased strategic fit according to claim 1 is characterized by: The construction process of the project portfolio selection model that matches the phased strategy in step S2 is as follows: S21: For the generality of model construction and analysis, the following assumptions are made: H1: The unit cycle is a standard time unit, including years, quarters, months, weeks or their integer multiples. The long-term strategic goals are implemented in phases according to T unit cycles, forming T+1 time nodes. At the initial point, the enterprise only launches new projects without terminating any projects. At the end point, the project portfolio enters the termination state and no new projects are launched at this time. H2: Each unit period is continuous, where t = 1 indicates the beginning of the first unit period, t = 2 indicates both the beginning of the second unit period and the end of the first strategic period, and so on; H3: If the projects currently being implemented can meet the needs of the strategic goals, no new projects will be selected for implementation. Projects that are not selected for implementation will enter the candidate project pool and their implementation possibility will be considered again after entering the next unit cycle. All projects in the project portfolio must be completed within the entire strategic cycle. H4: In the multi-period selection process of project portfolio, enterprise resources are allocated according to unit periods. To fit the reality, it is assumed that the amount of resources allocated in each unit period is flexible and variable, including leasing and purchasing, and the remaining resources that are not fully utilized in each unit period are automatically accumulated to the next strategic period. S22: Model constraint construction, specifically: S221: Regarding the execution constraints of the project's flexible execution strategy, specifically: The project is completed in a certain proportion but not all in a specific time period. The range of project execution ratio is [α it ,β it ], then the completion ratio of the project in the unit strategic cycle satisfies: pr t,j ∈0∪[α it ,β it ], the multi-phase flexible execution constraints of the project are: α t,j r t,j ≤pr t,j ≤β t,j r t,j ； Among them, j represents a certain project; S222: In the multi-period selection process of the water environment governance project portfolio, when the implementation requirements of the selected project portfolio exceed the limitations of the enterprise's own resources, the lead enterprise adopts a flexible resource allocation strategy, that is, the lead enterprise will obtain additional required resources through market-oriented means such as leasing or purchasing. Flexible resources are further divided into renewable flexible resources and non-renewable flexible resources. Multi-period flexible resource constraints are as follows: ① Flexible non-renewable resource constraints: In the formula, Indicates that the enterprise has its own non-renewable resources. It refers to the flexible non-renewable resources that enterprises obtain through procurement and buyout; ② Flexible renewable resource constraints: In the formula, It means that the enterprise has its own renewable resources. It refers to the flexible and renewable resources that enterprises obtain through deployment and leasing; S223: Logical constraints on project execution: Project execution needs to satisfy certain logical relationships. First, define the logical relationship set E between projects. loc , if the context relationship between two items (i,j)∈E loc , it means that project j can only start after project i is completed. The logical constraints of project execution are expressed as: y i1 =0, (i,j)∈E loc ; S224: Project completion constraints: ∑ t∈T ∑ j∈J r t,j ＝x j ； The project completion constraint states that if a project is selected, it must be completed within the entire investment period; S225: Project executable quantity constraints: Due to the limitations of resources, funds, and company capabilities, the number of projects that can be selected and executed within the unit cycle and the entire strategic cycle is limited. Therefore, it is necessary to set clear limits on the number of selectable projects. The constraint on the number of projects within a unit cycle is expressed as: Where N is the maximum number of projects that can be carried out simultaneously during the entire strategic cycle; S226: Multi-period water environment governance effect opportunity constraints: In the formula This is the water environment management standard value required by the government; The water environment governance project portfolio implemented in each unit cycle must achieve the governance objectives set by the current public sector. The multi-period water environment governance effect constraint ensures that the implementation of the project portfolio can not only achieve the overall governance objectives, but also achieve the objectives of each stage. In actual scenarios, enterprises usually leave a certain margin considering the fluctuation of governance effects during the implementation of water environment comprehensive projects, that is, the governance effect of the projects actually implemented will be higher than the minimum limit required by the government. S227: 0-1 constraints on decision variables: The 0-1 constraints corresponding to the decision variables are shown below: S23: Model objective function construction, specifically: The phased strategic fit objective function and the profit objective function of the project portfolio multi-period selection model are shown below. In order to take these two objectives into account, the compromise satisfaction strategy is introduced to construct the objective function of the project portfolio multi-period selection model. The phased strategic fit objective and the profit objective are expressed as follows: In the formula, and ρ m They represent the usage cost of unit flexible non-renewable resources and unit flexible renewable resources, DC j,t represents the additional cost incurred by the flexible execution of project j in period t, Q i represents the Q value of the i-th phase strategic target calculated by the GC-W / VIKOR method, is the priority weight of the project, FR t,i Represents the phased project portfolio selection scheme r t,j The achievement rate of the i-th phased strategic goal in period t; Then, the compromise satisfaction aggregation strategy is used to combine the phased strategy fit objective and the profit objective function into a single objective function. The specific process is as follows: S231: Calculate the optimal value and the worst value of each objective respectively, expressed as: and The optimal and worst values ​​of the target value are regarded as the left and right boundaries of the satisfactory solution interval. For each target, its closeness to the optimal value is μ i It is expressed as: S232: The improved Zimmerman method is used to aggregate the closeness of the optimal values ​​of the two objectives. The model using the compromise satisfaction aggregation strategy is expressed as follows: Where: W is the objective function value corresponding to the feasible solution that satisfies the model constraints, λ i (i∈{c,s}) is the psychological attention index, λ i The size of the reflection manager's attention mentality, i The closer it is to 1, the less attention is paid to this target value, and vice versa. SAT is the compromise satisfaction value of weighing the two goals. The multi-period model of water environment governance project portfolio for phased strategic fit is expressed as follows:

4. The multi-phase selection method for water environment management project portfolio with phased strategic fit according to claim 1 is characterized by: The solution process of the simulated runoff mimicry algorithm in step S3 is: S31: The runoff algorithm first randomly generates a number of water flows to form a population. The calculation method of the initial position of the runoff is: L X(i) =x L +r d ×(x H -x L ); Where, L X(i) represents the location of the runoff at the i-th location, x H and x L represents the upper and lower limits of the initial runoff position, r d is a random value uniformly distributed between 0 and 1. Then, it is assumed that there are J neighbors near each runoff location, and the location of the neighborhood is expressed as: N X(j) =L X(i) +r dn ×Δ; Where N X(j) represents the position of the jth neighbor (j∈J), r dn is a random value that obeys the normal distribution, Δ is the preset search space parameter. The larger the Δ value, the larger the search space of the runoff algorithm and the stronger the global optimization ability. The smaller the Δ value, the higher the accuracy of the local optimization. Therefore, determining the appropriate Δ value is the key to ensuring the performance of the algorithm. Otherwise, only pursuing global optimization will make it difficult for the algorithm to accurately locate the optimal value, while only pursuing local optimization will easily cause the algorithm to fall into the local optimal solution. When using the linear random strategy, the calculation method of the Δ value in the runoff algorithm is as follows: In the formula, By L X(i) The random position generated, the first term represents L X(i) Gradually move to a new random position The second term represents L X(i) Gradually move to the optimal position B through iteration X , so that B X and L X(i) The Euclidean distance of ||B X -L X(i) || gradually approaches 0, which makes the algorithm search space gradually shrink and tend to be stable. rd is a random weight strategy, calculated as: In the formula, is a uniformly distributed random vector, and W is a random weight strategy, which further introduces randomness into the runoff algorithm, thereby preventing the algorithm from falling into the local optimum; In the runoff algorithm, the runoff velocity V is positively correlated with the slope of its location, so the flow velocity V is expressed as: V=r dn ×g0(i,j,d); Where g0 represents the slope of the runoff location, which is determined by: In the formula, F(·) represents the fitness value (altitude) of the location where the runoff is calculated, and the new location of the runoff is The update formula is expressed as: Then by iterating the runoff position Directing runoff to lower elevations Thus, the final outlet is found, which means that the runoff algorithm finds the global optimal solution; S32: Use the Tent chaos strategy to generate the initial population, specifically: Tent mapping is used to generate chaotic sequences so that the initial solutions are distributed as evenly as possible in the solution space. This aims to enhance the algorithm's optimization ability by improving the uniformity of the initial solutions. The method for generating chaotic sequences based on Tent mapping is as follows: In the formula, N p is the population number, I t is the current iteration number, according to the chaotic sequence Generate a sequence of runoff locations within the search area The process is as follows: In the formula, and They are The maximum and minimum values ​​of the sequence; S33: Update the neighborhood flow position using the sinusoidal strategy, specifically: In the runoff algorithm, since the search space operator Δ is a linear transformation in a random direction, the optimization efficiency of the algorithm is limited. Based on this, the sinusoidal strategy is used to improve the operator Δ and the neighborhood flow position update method is adjusted, so as to better balance the global search ability and local optimization ability of the algorithm on the basis of ensuring the search space. The improved runoff neighborhood position update method and search space operator Δ * The calculation formula is as follows: N X(i) =L X(i) ×Δ * ; Where R1 and R2 are random angle numbers that determine the displacement distance and displacement direction in the next iteration, and σ1 and σ2 are based on the golden section coefficients. The obtained coefficients.

Images