Multi-objective optimization design method for current collection system of water surface photovoltaic power station

Abstract

The invention relates to a multi-objective optimization design method for a current collection system of a water surface photovoltaic power station, and solves the problems that a manual design scheme is low in automation degree, poor in economy and time-consuming in design, and a sequential optimization scheme is easy to fall into a local optimal solution. The method comprises the following steps: establishing a power generation unit layout cost model and an annual energy production model associated with an optimized variable capacity ratio based on the definition of a minimum power generation unit, then integrating an improved FCM class algorithm, a DM-MSTP algorithm and automatic selection and model selection of a current collection cable, and establishing a current collection cable laying cost model; the method comprises the following steps: establishing a water surface photovoltaic power station current collection system design model, setting constraint conditions required to be met by the water surface photovoltaic power station current collection system design, establishing a water surface photovoltaic power station current collection system multi-target optimization design model, integrating a non-dominated sorting artificial collaborative search algorithm and Pareto evaluation, and generating a multi-target optimization design scheme of an optimal solution based on a combined empowerment TOPSIS method. Therefore, the purpose of optimal design of the water surface photovoltaic power station current collection system is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of floating photovoltaic power stations, and particularly to a multi-objective optimization design method for the collector system of a floating photovoltaic power station. Background Technique

[0002] With the development of solar photovoltaics, the available land for further development of photovoltaic power stations has decreased. Due to the potential advantages, expected profits of floating photovoltaic power stations, and the lack of land resources, more and more floating photovoltaic power stations have obtained investment and construction and achieved grid-connected power generation.

[0003] The optimization design of the collector system of a floating photovoltaic power station is an important link during the construction of a floating photovoltaic power station, mainly including the layout optimization of photovoltaic power generation units (referred to as power generation units for short) and the laying optimization of collector cables. Currently, the design of the collector system of a floating photovoltaic power station is still based on manual experience, with technical deficiencies such as low automation, poor economy, and time-consuming design.

[0004] In view of the current situation that the design of the collector system of a floating photovoltaic power station is still based on manual experience, for this reason, the research results of the optimization design of the collector system of an offshore wind farm are used for reference. For the planning and design of an offshore wind farm, a sequential optimization scheme is usually adopted, first optimizing the layout of wind turbines and then optimizing the laying of collector cables. However, the sequential optimization scheme may lead to the problem of suboptimal solutions in the planning and design of an offshore wind farm. The reason is that during the layout of wind turbines, the distance between wind turbines is increased to reduce the influence of wake effects, and the implementation of the sequential optimization scheme will lead to a reduction in the number of wind turbines and an increase in the length of collector cables, thereby resulting in a reduction in power generation and an increase in the investment cost of collector cables. Similarly, for a floating photovoltaic power station, increasing the capacity ratio during the layout of power generation units to increase power generation will lead to an increase in the laying cost of photovoltaic panels and the investment cost of collector cables. In addition, for offshore wind farms and floating photovoltaic power stations, the laying cost and the investment cost of collector cables are both one-time expenditures, occurring during the construction period of new energy power stations, while power generation is a long-term benefit, running through the entire operation period of new energy power stations. In terms of the overall benefit of new energy power stations, the costs and benefits during their construction period and operation period should be considered as a whole to achieve the coordinated optimization of power generation and economic costs.

[0005] Chinese patent document CN119558014A discloses a method for collaborative optimization design of the layout of power generation units and collector cables in a floating photovoltaic power station. This method proposes the layout optimization of the power generation units in the collector system of a floating photovoltaic power station based on the minimum power generation unit, as well as the optimization of collector cables integrating fuzzy C-means clustering, DM-MSTP algorithm and automatic selection of collector cables. Then, the particle swarm algorithm is used to conduct collaborative optimization of the layout of power generation units and collector cables in the floating photovoltaic power station with the total investment cost of the floating photovoltaic power station as the target, so as to achieve the purpose of collaborative optimization design of the layout of power generation units and collector cables in the collector system of the floating photovoltaic power station. However, this solution only considers the single target of the total investment cost of the floating photovoltaic power station and fails to well coordinate the relationship between the practicability and economy of the collector system of the floating photovoltaic power station. At the same time, compared with multi-objective optimization, single-objective optimization has great limitations.

[0006] Therefore, the problem of multi-objective optimization design of the collector system of a floating photovoltaic power station has become a key technical problem that urgently needs to be solved in the technical field of floating photovoltaic power stations. Summary of the Invention

[0007] In order to solve the deficiencies of low automation, poor economy and time-consuming design in the existing manual design scheme, as well as the deficiency that the sequential optimization scheme is prone to fall into local optimal solutions, the present invention provides a method for multi-objective optimization design of the collector system of a floating photovoltaic power station.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] A method for multi-objective optimization design of the collector system of a floating photovoltaic power station, the method comprising the following steps:

[0010] S1. Obtain the basic data of the floating photovoltaic power station.

[0011] S2. Establish a multi-objective optimization design model for the collector system of the floating photovoltaic power station.

[0012] S3. Use the multi-objective optimization design model for the collector system of the floating photovoltaic power station to conduct multi-objective optimization to obtain a multi-objective optimization design scheme.

[0013] S4. Based on the basic data of the floating photovoltaic power station and the multi-objective optimization design scheme, use the CWM-TOPSIS method to obtain the design result of the collector system of the floating photovoltaic power station.

[0014] In steps S3 and S4, based on the obtained basic data, NSACS and Pareto evaluation are integrated to perform multi-objective optimization on the established mathematical model for the design of the collector system of the floating PV power station, obtaining the Pareto front. Then, based on the CWM-TOPSIS method, the optimal design scheme for the collector system of the floating PV power station is obtained from the Pareto front, and a design drawing of the collector system of the floating PV power station is generated.

[0015] As a further improvement of the above technical solution, in step S1, the basic data of the floating PV power station includes: light resource and water area data and device basic data; the light resource and water area data includes the light resource level, annual sunlight conditions, and water area conditions of the location area of the floating PV power station; the device basic data includes the model and pricing of the inverter, the model and pricing of the PV panels, and the model of the collector cable.

[0016] As a further improvement of the above technical solution, step S2 specifically includes the following steps:

[0017] S21. Based on the minimum power generation unit, establish a power generation unit layout cost model.

[0018] S22. Establish an annual power generation model associated with the optimization variable capacity ratio of the collector system of the floating PV power station.

[0019] S23. Based on the improved FCM clustering algorithm, DM-MSTP algorithm, and the automatic selection and sizing method of collector cables, establish a collector cable laying cost model.

[0020] S24. Set the constraint conditions that need to be satisfied for the design of the collector system of the floating PV power station.

[0021] S25. According to the power generation unit layout cost model, the annual power generation model, the collector cable laying cost model, and the constraint conditions, establish a multi-objective optimization design model for the collector system of the floating PV power station.

[0022] As a further improvement of the above technical solution, step S3 specifically includes the following steps:

[0023] S31. Combine the NSACS algorithm and Pareto evaluation to generate the Pareto solution set front.

[0024] S32. Based on the CWM-TOPSIS method, generate a multi-objective optimization design scheme for the optimal solution from the Pareto solution set front.

[0025] As a further improvement of the above technical solution, step S21 specifically includes the following steps:

[0026] S211. Define a power generation unit composed of one inverter and n photovoltaic modules as the minimum power generation unit of the floating photovoltaic power station. Based on the definition of the minimum photovoltaic power generation unit, calculate the capacity of the minimum power generation unit and the number of photovoltaic modules in the minimum power generation unit by using Equation (1) and Equation (2) respectively:

[0027] P min = P INV ×λ (30)

[0028] In Equation (1), P min is the capacity of the minimum power generation unit, P INV is the rated power of the inverter, and λ is the capacity ratio.

[0029]

[0030] In Equation (2), n is the number of photovoltaic modules in the minimum power generation unit; P min is the capacity of the minimum power generation unit; P gfb is the rated power of the photovoltaic panel; N zj is the number of photovoltaic panels in a photovoltaic module, and this number is restricted by the floating body structure and should be selected from fixed specifications; n is an integer and is rounded by using the rounding function.

[0031] S212. Calculate the capacity of the power generation unit by using Equation (3):

[0032] K a ×K b = n × N p.unit (32)

[0033] In Equation (3), the power generation unit is the basic component unit of the floating photovoltaic power station and is composed of N p.unit minimum power generation units spliced together. Arrange the photovoltaic modules in the power generation unit as a rows and b columns. Assume that there are K b photovoltaic modules in each row and K a photovoltaic modules in each column; N p.unit is the number of minimum power generation units.

[0034] S213. Calculate the length and width of the minimum power generation unit in the floating photovoltaic power station by using Equation (4) and Equation (5):

[0035] L = L unit ×K b (33)

[0036] W = W unit ×K a (34)

[0037] In Equation (4) and Equation (5), L and W respectively represent the length and width of the minimum power generation unit in the floating photovoltaic power station, Lunit , W unit respectively represent the length and width of each photovoltaic module in the minimum power generation unit, and the length and width of each photovoltaic module are determined by the selected photovoltaic panel model, photovoltaic inclination angle, and the spacing between panels; K b represents the number of photovoltaic modules in each row of the power generation unit; K a represents the number of photovoltaic modules in each column of the power generation unit.

[0038] S214. Set the capacity ratio of the floating photovoltaic power station, the number of minimum power generation units, and the aspect ratio of the power generation unit as optimization variables. Under the condition of ensuring that the power generation unit can be laid in the target water area, set the layout cost of the power generation unit of the floating photovoltaic power station as the objective function f1. Then, use Equation (6) to determine the objective function f1, and further determine the layout cost model of the power generation unit according to the objective function f1:

[0039] f1 = (C INV + C gfb × N zj × n) × N p.unit × N unit (35)

[0040] Substitute Equation (1) and Equation (2) into Equation (6), then we have:

[0041]

[0042] In Equation (6) and Equation (7), f1 is the objective function of the layout cost of the power generation unit, C INV is the cost of a single inverter, C gfb is the cost of a single photovoltaic panel, N zj is the number of photovoltaic panels in a photovoltaic module, n is the number of photovoltaic modules in the minimum power generation unit, N p.unit is the number of minimum power generation units, N unit is the number of power generation units of the floating photovoltaic power station, P INV is the rated power of the inverter, λ is the capacity ratio, P gfb is the rated power of the photovoltaic panel.

[0043] For the collector system of a floating PV power station, the ratio of the nominal power of the PV modules on the DC side to the output power of the PV inverter is called the capacity ratio. Increasing the capacity ratio is achieved by increasing the number of PV modules and over-sizing the DC side capacity of the PV system. Although increasing the capacity ratio increases the initial investment cost of the system to a certain extent, in terms of the entire life cycle of PV power generation, increasing the capacity ratio can effectively improve the utilization rate of the inverter capacity, and thus increase the PV power generation. According to the objective function f1 of the layout cost of the power generation units in the floating PV power station, the layout cost model of the power generation units is determined, and the layout cost of the power generation units is changed by adjusting and optimizing variables (such as the capacity ratio, the number of the smallest power generation units, and the aspect ratio of the power generation units, etc.).

[0044] As a further improvement of the above technical solution, the step S22 specifically includes the following steps:

[0045] S221. Calculate the total PV power generation of the floating PV station by using Equation (8):

[0046]

[0047] In Equation (8), P out is the total PV power generation within the unit time stamp; PR is the energy efficiency ratio; P0 is the standard installed capacity, that is, the installed capacity under the 1:1 capacity ratio; G i,k is the cumulative irradiation amount on the inclined plane within the unit time stamp, with the unit of kWh / m 2 ; τ k is the unit time stamp; G i,ref is the standard irradiance, usually taken as 1000W / m 2 .

[0048] Equation (8) is the power generation calculation formula in the international standard IEC 61724-1:2017 Photovoltaic Performance Assessment Standard. It can be seen from Equation (8) that the power generation formula in the international standard is affected by various factors, and there is a strong coupling between the influencing factors.

[0049] S222. Decouple the power generation formula (8), establish the relationship between the power generation and the capacity ratio, and set the negative value of the annual power generation of the floating PV station as the objective function f2. Based on the objective function f2, establish an annual power generation model associated with the optimization variable capacity ratio of the collector system of the floating PV station:

[0050] f2 = -P out = -PR0 × PR1 × T ref × P o × λ × [1 - a - b × (y - 1)] (38)

[0051] P0 = P INV × N p.unit × Nunit (39)

[0052] In Formulas (9) and (10), f2 is the objective function of the annual power generation of the floating PV power station; P out is the total PV power generation within a unit time stamp; PR0 is the fixed loss rate on the DC side, representing the DC side losses other than the light rejection rate and the component attenuation rate; PR1 is the AC side conversion efficiency; T ref is the local annual peak hours, which can be queried through the historical meteorological database; P0 is the standard installed capacity, that is, the installed capacity under a 1:1 capacity ratio; λ is the capacity ratio; a is the first-year attenuation rate of the component; b is the second-year linear attenuation rate of the component; y is the number of years of system operation. Generally, the life cycle of PV modules is 25 years; P INV is the rated power of the inverter; N p.unit is the number of minimum power generation units; N unit is the number of power generation units of the floating PV power station.

[0053] As a further improvement of the above technical solution, S23 specifically includes the following steps:

[0054] S231. Set the collector cable model as an optimization variable and establish constraint conditions using Formula (11):

[0055]

[0056] In Formula (11), S q·f is the capacity of the f-th box-type substation towed by the q-th collector cable, S q·max represents the maximum apparent power of the q-th collector cable, F represents the number of box-type substations in the q-th collector cable, Q is the number of collector cables in the collection system, I represents the current-carrying capacity of the collector cable, k represents the number of towed box-type substations, P N is the rated power of the power generation unit of the floating PV power station, U is the collection system voltage, cosθ is the power factor of the PV panel, N max is the maximum number of box-type substations that a single collector cable can bear, I max is the maximum current-carrying capacity of the collector cable; the constraint conditions include the collector cable capacity constraint, the current-carrying capacity constraint, and the constraint on the maximum number of towable box-type substations; P N represents the rated power of the power generation unit of the floating PV power station.

[0057] To overall reduce the cost of collector cable laying, the present invention sets the collector cable model as an optimization variable. During the selection process, it should meet the collector cable capacity constraint, the current-carrying capacity constraint, and the constraint on the maximum number of towable box-type substations.

[0058] S232: Taking the onshore booster station as the radiation center, use the improved FCM clustering algorithm to partition the box-type substations. Based on the above-mentioned constraints, use Equation (12) to calculate the distances between the box-type substations in each partition and between the box-type substations and the onshore booster station, and construct a distance matrix.

[0059]

[0060] In Equation (12), L d is the element in the distance matrix of the d-th partition; are the abscissa and ordinate of the box-type substations in the d-th partition respectively, where d = 1, 2…, N, i = 1, 2…, m.

[0061] According to the constraints in (11), considering that the floating PV power station has a large power generation capacity and many box-type substations, it is impossible for the collector cables to carry all the box-type substations at the same time. Therefore, the box-type substations are partitioned first, and then the collector cables are laid. The present invention takes the onshore booster station as the radiation center and uses the improved FCM clustering to partition the box-type substations. Since the more the number of partitions, the fewer the number of box-type substations in each area, which will result in a longer length of the required collector cables. Therefore, the initial number of partitions is set to 1 and incremented successively until the maximum number of box-type substations that can be carried by a single collector cable is satisfied, and then stop. At this time, it is the minimum number of partitions that satisfies the current-carrying capacity constraint, and this number is denoted as N.

[0062] Based on the research of using the improved FCM clustering algorithm to realize the partition of box-type substations, taking the onshore booster station as the radiation center, calculate the distances between the box-type substations in each partition and between the box-type substations and the onshore booster station, and construct a distance matrix. Since the wiring requirements of the floating PV power station are "horizontal and vertical", the distance calculation formula is expressed as Equation (12).

[0063] S233: Based on the above-mentioned constraints and the distance matrix, use the DM-MSTP algorithm to optimize the length of the collector cables of the floating PV power station, then use the collector cable automatic type selection algorithm to optimize the type selection of the collector cables, determine the laying scheme of the collector cables of the floating PV power station, and finally establish a collector cable laying cost model based on the laying scheme of the collector cables of the floating PV power station.

[0064] As a new minimum spanning tree solution method, the DM-MSTP algorithm collects data through an adjacency matrix. Compared with the Prim algorithm, which is a solution to the classical minimum spanning tree problem, the DM-MSTP algorithm has the advantage of low time complexity. The present invention optimizes the length of the collector cables of the floating PV power station based on the DM-MSTP algorithm, and then takes the two connection methods of tree connection and chain connection as examples to optimize the type selection of the collector cables through the proposed collector cable automatic type selection algorithm.

[0065] The S233 specifically includes the following steps:

[0066] S2331. Based on the constraint conditions and the distance matrix, use the DM-MSTP algorithm to optimize the length of the integrated cable of the floating PV power station, and generate a collector cable connection table for each partition; the first row of the inherited cable connection table represents the starting node, the second row represents the ending node, and the third row represents the distance between the starting node and the ending node.

[0067] S2332. Based on the integrated cable connection table, calculate the number of times each node appears. If the number of times all nodes appear is less than or equal to 2, record the number of box-type substations mounted on each node, and perform collector cable type selection according to the constraint conditions of Equation (11), and execute step S2335; if there is a node among all nodes whose number of appearances is greater than 2, execute step S2333.

[0068] S2333. In the collector cable connection table, find the node with an appearance count of 1 other than the step-up substation on the road, record the position of this node in the corresponding table as s, and let r = s - 1; compare the values of Link(1, s) and Link(2, r), where Link(1, s) represents the element in the first row and the s-th column of the table, and Link(2, r) represents the element in the second row and the r-th column; if Link(1, s) = Link(2, r), record the number of box-type substations mounted on the node Link(2, r), then subtract 1 from the value of s and subtract 1 from the value of r, and continue to record the corresponding number of box-type substations mounted; if Link(1, s) ≠ Link(2, r), then subtract 1 from the value of r, and continue to determine whether Link(1, s) is equal to Link(2, r - 1), and traverse forward in this way until the serial number is 1, and end the traversal process; record all the above connection processes in the same cell array, and denote this array as LJ. After completion, transfer to step S2334.

[0069] S2334. In the collector cable connection table, find the node with an appearance count greater than 2, and record the found node as u; for the cell array LJ, merge the arrays containing u; the merging rule is to first determine the position of u in the cell array, and then compare the number of box-type substations mounted on the u node in different arrays containing u, and add the number of box-type substations mounted on the u node with the smaller number to the u node with the larger number; continue to merge according to this rule until there is only one cell left in the cell array LJ. At this time, according to the constraint conditions specified in Equation (11), use the collector cable automatic type selection algorithm to optimize the type selection of the collector cable.

[0070] S2335. Based on the selection of the collector cable, that is, the selection of the chain-shaped and tree-shaped collector cables, determine the laying scheme of the collector cable for the floating photovoltaic power station, and then establish a collector cable laying cost model using Equation (13).

[0071]

[0072] In Equation (13), f3 is the objective function of the collector cable laying cost, N is the number of partitions of the fuzzy C-clustering, m is the number of box-type substations in each partition, L C.q is the length of the q-th collector cable, and C C.q (φ) is the cost per kilometer of the q-th collector cable, and φ is the cross-sectional area of the collector cable.

[0073] In the present invention, the collector cable model is set as the optimization variable. Through the selection of the collector cable model, the partitioning of the box-type substation, and the optimization of the collector cable length and type selection to achieve the optimization of the collector cable laying, and then the collector cable laying cost is set as the objective function f3.

[0074] As a further improvement of the above technical solution, in step S25, the multi-objective optimization design model of the collector system of the floating photovoltaic power station is shown in Equation (14).

[0075] Set the capacity ratio, the minimum number of power generation units, the aspect ratio of the power generation unit, and the collector cable model as the optimization variables, and the layout cost of the power generation units, the annual power generation, and the collector cable laying cost of the floating photovoltaic power station as the objective functions. Considering that all power generation units cannot exceed the water area boundary, there is no intersection or overlap between the power generation units, and the width reserved for the operation and maintenance channel should meet the national standard requirements. At the same time, considering that the collector cable needs to meet its capacity limit condition, that is, the maximum load-bearing capacity of the selected collector cable model must be greater than or equal to the sum of the capacities of the box-type substations it drives downstream, establish a multi-objective optimization model of the collector system of the floating photovoltaic power station considering the constraint conditions, as shown in Equation (14).

[0076] min(f1, f2, f3)

[0077]

[0078] In Equation (14), x i,v , y i,v are the horizontal and vertical coordinates of each vertex of the power generation unit respectively, where i = 1, 2... m represents the power generation unit number, and v = 1, 2, 3, 4 represents the vertex number of the power generation unit; x min , x max , y min , y max represent the boundary coordinates of the water area respectively, L space is the width of the operation and maintenance channel, and S q·fis the capacity of the f - th box - type substation towed by the q - th collector cable, S q·max represents the maximum apparent power of the q - th collector cable, F represents the number of box - type substations in the q - th collector cable, Q is the number of collector cables in the collector system, I represents the current - carrying capacity of the collector cable, k represents the number of towed box - type substations, P N is the rated power of the power generation unit of the floating PV power station, U is the voltage of the collector system, cosθ is the power factor of the PV panel, N max is the maximum number of box - type substations that a single collector cable can bear, I max is the maximum current - carrying capacity of the collector cable.

[0079] As a further improvement of the above - mentioned technical solution, the artificial collaborative search algorithm is an advanced optimization algorithm that solves complex optimization problems by simulating the cooperation and information - sharing mechanism among multiple agents. In order to achieve the multi - objective optimization of the collector system of the floating PV power station, first, the NSACS algorithm and the Pareto evaluation are combined to generate the Pareto solution set frontier. The Pareto solution set frontier is used to achieve the multi - objective optimization of the design optimization model of the collector system of the floating PV power station.

[0080] The step S31 specifically includes the following steps:

[0081] S311. Initialize the population and conduct fitness evaluation

[0082] S3111. Use Equation (15) to create the initial populations for two super - individuals A and B respectively, and calculate the fitness (f1(x), f2(x)....f n (x)) of each individual in the super - individuals A and B.

[0083] Specifically, the population in each super - individual is composed of multiple individuals, which are the basic units for subsequent optimization calculations. After creating the initial population, for each individual in the super - individuals A and B, calculate its fitness (f1(x), f2(x)....f n (x)). Here, the fitness is a key indicator to measure the performance of an individual in the multi - objective optimization problem. Among them, f1(x), f2(x) up to f n (x) represent multiple objective function values related to the collector system of the floating PV power station, including objective functions related to the layout cost of the power generation unit, annual power generation, the laying cost of the collector cable, etc. The fitness values calculated through these functions can evaluate the advantages and disadvantages of each individual in achieving the multi - objective optimization of the system.

[0084] S3112. Sort each individual in A and B using the fitness function values, and calculate the crowding distance D of each individual using Equation (16) c, then store the fitness, ranking, and CD of each individual in the A and B super-individuals;

[0085]

[0086] In equations (15) and (16), p = 1...M, q = 1...D, M is the population size, D is the problem dimension, rand is a random number uniformly distributed, low q and up q are the search lower and upper bounds of q respectively, y p,A and y p,B are the fitnesses generated by the two super-individuals A and B respectively, fitnessA and fitnessB are the fitnesses of A and B respectively, D c (A1) is the crowding degree of individual A1, A2 and A3 are two individuals adjacent to A1, f l (A2) and f l (A3) are the objective function values of individuals A2 and A3 respectively, A0 is the neighborhood center of individuals A2 and A3, f l (A0) is the objective function value of the neighborhood center.

[0087] S312, Selection of Predator and Prey

[0088] For the super-individuals A and B in each parallel computing area, use equation (17) to randomly select the predator and prey:

[0089]

[0090] In equation (17), a and b are randomly generated numbers, predator represents the predator, prey represents the prey, := represents the update operation, y predat or, y prey 、y A 、y B represent the fitnesses generated corresponding to predator, prey, A, and B respectively.

[0091] S313, Offspring Generation

[0092] S3131, Utilize the biological interaction between the predator and the prey to generate a new position and direction for the predator, and at the same time implement boundary control. Implementing boundary control can ensure that the predator is within a reasonable search range. The change in position and direction can be completed using the Wiener process. Predators obtain stronger hunting ability through selection crossover.

[0093] S3132, Predators perform selection crossover to enhance their hunting ability. Based on the hunting ability, use equation (18) to determine the probability of cooperation between predators.

[0094] S3133. Dynamically adjust the dynamic factor of the search position using Equation (19) to generate offspring of the predator.

[0095] The dynamic factor includes the mutation rate and the selection crossover rate of the search position. By dynamically adjusting the dynamic factor of the search position, offspring of the predator with stronger hunting ability are generated, and the offspring continue the search and hunting.

[0096]

[0097] In Equations (18) and (19), x is the mutation matrix, predator represents the predator, prey represents the prey, a, b, and c are randomly generated numbers, P is the control parameter, x p,q 、stronger p,q 、predator p,q are the mutation matrix value, the stronger predator, and the predator at the corresponding (p, q); the dot represents the dot product operation of the matrix or vector.

[0098] S3134. Calculate and store the offspring information. Calculate the fitness, rank, and crowding distance CD of the offspring, and store this information in a separate location.

[0099] S314. Elite selection

[0100] Find the optimal population by combining the predator and the offspring, sort the combined population according to the fitness and rank, and select the optimal individuals.

[0101] S315. Repeat steps S311 - S314 until the maximum number of iterations to generate the Pareto solution set frontier for the multi-objective optimization of the collector system of the floating PV power station.

[0102] As a further improvement of the above technical solution, based on the generated Pareto solution set frontier, use the combined weighted TOPSIS to determine the optimal design scheme of the collector system of the floating PV power station.

[0103] The step S32 specifically includes the following steps:

[0104] S321. Based on the subjective weight and the objective weight Use Equation (20) to obtain the combined weight ω * :

[0105]

[0106] In Equation (20), ω * is the combined weight, α1 and α2 are the weight coefficients, is the subjective weight, is the objective weight.

[0107] S322. Use the concept of game theory shown in Equation (21) to find the Nash equilibrium point:

[0108] Min||ω * -ω k ||2, k = 1, 2 (50)

[0109] In Equation (21), ω * is the combined weight, ω k represents the transpose of the weight, where k = 1, 2.

[0110] S323. According to the properties of matrix differentiation, the optimal solution of Equation (21) satisfies the conditions shown in Equation (22):

[0111]

[0112] In Equation (22), is the subjective weight, is the objective weight, α1 and α2 are weight coefficients, ω1 and ω2 are the transposes of the subjective weight and the objective weight.

[0113] S324. After normalization, obtain and calculate the optimized combined weight using Equation (23):

[0114]

[0115] In Equation (23), ω * is the combined weight, and are the normalized weight coefficients, is the subjective weight, is the objective weight.

[0116] S325. Use TOPSIS to evaluate the generated Pareto front. After obtaining the optimized combined weight, use TOPSIS to evaluate the front of the generated Pareto solution set.

[0117] In the above S325, the evaluation of the front of the generated Pareto solution set using TOPSIS includes:

[0118] S3251. Assume there are m objects and n evaluation criteria, and construct a decision matrix.

[0119] S3252. Normalize and weight the decision matrix to obtain the weighted decision matrix V shown in Equation (24):

[0120]

[0121] In Equation (24), V is the weighted decision matrix, and b ij is the corresponding weight, where i = 1...m and j = 1...n.

[0122] S3253. Evaluate the generated Pareto solution set front using the weighted decision matrix V.

[0123] S326. Determine the positive ideal solution F + and the negative ideal solution F - , and calculate the best distance and the worst distance of each solution using Equations (27) and (28):

[0124] F + = max 1≤i≤m v ij (54)

[0125] F - = min 1≤i≤m v ij (55)

[0126]

[0127] In Equations (25) to (28), F + and F - are the positive ideal solution and the negative ideal solution respectively, v ij is the value at (i, j) in the weighted decision matrix, and are the best distance and the worst distance respectively, F i + and F i - are the positive ideal solution and the negative ideal solution of the i-th row in the weighted decision matrix V respectively, and Fi j is the ideal solution value at (i, j).

[0128] S327. Calculate the comprehensive evaluation index C j using Equation (29), and sort the solutions according to the comprehensive evaluation index C j to determine the optimal solution of the Pareto front, and then determine the optimal design scheme of the collector system of the floating PV power station:

[0129]

[0130] In Equation (29), C j is the comprehensive evaluation index, and are the best distance and the worst distance respectively.

[0131] As a further improvement of the above technical solution, step S4 specifically includes the following steps:

[0132] S41. Based on the available water area, water area restricted areas, and operation and maintenance channel requirements, obtain the layable area of the power generation units of the floating PV power station, and based on national standard requirements and design resources, obtain the PV panel model, inverter model, collector cable group, and number of associated components.

[0133] S42. Initialize the capacity ratio λ, minimum number of stacked power generation units N p.unit of the floating PV power station, the aspect ratio K of the power generation unit, and the collector cable model, and set the number of initialized super individuals to 200 and the number of iterations to 30 times.

[0134] S43. According to the capacity ratio λ, minimum number of power generation units N p.unit in the super individual, the aspect ratio K = L / W value of the floating PV power generation unit, and the constraint conditions in Equation (13), lay the power generation units in the water area space: If the values of λ, N p.unit and K = L / W in this group of super individuals cannot achieve the laying of power generation units, directly jump to the next group of super individuals; if the laying of power generation units can be achieved, calculate the objective function f1 using Equation (7), calculate the objective function f2 using Equation (9), and generate the coordinates of the box-type substation according to the setting that the box-type substation is arranged at the center of the power generation unit; according to the collector cable model in this group of super individuals, based on the constraint conditions in Equation (13), use the improved FCM algorithm to partition the box-type substation with the roadside step-up substation as the base point, and then integrate the DM-MSTP algorithm and the proposed collector cable automatic selection algorithm to lay the collector cables; if the collector cables in this group of super individuals are not sufficient to tow the box-type substation in any partition, jump to the next group of super individuals; if all box-type substations in all partitions can be towed, then calculate the objective function f3 according to Equation (12).

[0135] S44. Terminate the iteration when the number of iterations is reached, and generate the Pareto front of the multi-objective optimization of the collector system of the floating PV power station.

[0136] S45. Based on the CWM-TOPSIS method, output the optimal design scheme of the collector system of the floating PV power station from the Pareto solution set front of the multi-objective optimization of the collector system of the floating PV power station generated, and obtain the design results of the collector system of the floating PV power station.

[0137] Compared with the prior art, the advantages of the present invention are:

[0138] In terms of the design of the collector system of a floating PV power station, the existing manual design scheme has the deficiencies of low automation, poor economy, and time-consuming design, while the sequential optimization scheme is prone to falling into suboptimal solutions. Therefore, in order to overall consider the layout cost of the power generation units, the annual power generation, and the laying cost of the collector cables of the floating PV power station collector system, and to achieve a reasonable design of the floating PV power station collector system, the present invention proposes a multi-objective optimization design method for the floating PV power station collector system. This method first establishes a power generation unit layout cost model based on the definition of the minimum power generation unit, and then establishes an annual power generation model directly related to the capacity ratio; then, it integrates the improved FCM clustering algorithm, the DM-MSTP algorithm, and the automatic selection and sizing of collector cables, establishes a collector cable laying cost model, and reasonably sets the comprehensive constraints required for the design of the floating PV power station collector system. Furthermore, a multi-objective optimization design model for the floating PV power station collector system is established, and the NSACS algorithm and the Pareto evaluation are integrated, and the optimal solution of the multi-objective optimization design scheme is generated based on the CWM-TOPSIS method. The method described in the present invention effectively takes into account the layout cost of the power generation units, the annual power generation, and the laying cost of the collector cables of the floating PV power station collector system on the basis of making full use of the water area space, realizes the optimized design of the floating PV power station collector system, and at the same time has the value of reference for actual engineering design and popularization and application. Description of the Drawings

[0139] Figure 1 is the method sequence diagram of the multi-objective optimization design method for the floating PV power station collector system in the present invention;

[0140] Figure 2 is the multi-objective optimization flowchart of the floating PV power station collector system in the present invention;

[0141] Figure 3 is the schematic diagram of the collector cable laying of Case1;

[0142] Figure 4 is the schematic diagram of the collector cable laying of Case2;

[0143] Figure 5 is the schematic diagram of the collector cable laying of Case3;

[0144] Figure 6 is the design diagram of the collector system of Case 4;

[0145] Figure 7 is the design diagram of the collector system of Case 5;

[0146] Figure 8 is the schematic diagram of the Pareto front corresponding to Case 6;

[0147] Figure 9It is the design drawing of the power collection system for Case 6. Specific implementation manners

[0148] To solve the deficiencies of the manual design scheme in the prior art, such as low automation level, poor economy, and time-consuming design, as well as the deficiency that the sequential optimization scheme is prone to falling into local optimal solutions, and the deficiency that Chinese Patent Document CN 119558014A only considers the single objective of the total investment cost of the floating photovoltaic power station and fails to well coordinate the relationship between the practicability and economy of the floating photovoltaic power station's power collection system, and the single-objective optimization has great limitations. For this reason, the present invention establishes an optimization design model for the floating photovoltaic power station's power collection system with the layout cost of the power generation units, the annual power generation, and the laying cost of the power collection cables as the objective functions, and proposes to comprehensively use NSACS and Pareto evaluation to perform multi-objective optimization on the established mathematical model of the floating photovoltaic power station's power collection system design to obtain the Pareto front, and then obtain the optimal design scheme for the floating photovoltaic power station's power collection system from the Pareto front based on the CWM-TOPSIS method.

[0149] As Figure 1 shown, the present invention proposes a multi-objective optimization design method for the floating photovoltaic power station's power collection system. This method establishes a layout cost model for the power generation units based on the definition of the minimum power generation unit, and then establishes an annual power generation model associated with the variable capacity ratio of the optimization variables. Subsequently, an integrated laying cost model for the power collection cables is established by integrating the improved Fuzzy c-means (FCM) clustering algorithm, the DM-MSTP (Dhouib Matrix Minimum Spanning Tree Problem) algorithm, and the automatic selection and sizing of the power collection cables. Then, the constraint conditions required for the design of the floating photovoltaic power station's power collection system are set, and further a multi-objective optimization design model for the floating photovoltaic power station's power collection system is established. The multi-objective optimization process is as Figure 2 shown. The present invention proposes a multi-objective optimization design scheme that integrates the Non dominated sorting artificial collaborative search (NSACS) algorithm and Pareto evaluation based on non-dominated sorting, and then generates the optimal solution based on the Composite Weighted TOPSIS (CWM-TOPSIS) method generated by the TOPSIS method with combined weights, so as to achieve the purpose of optimizing the design of the floating photovoltaic power station's power collection system. Based on the obtained basic data, the present invention integrates NSACS and Pareto evaluation, performs multi-objective optimization on the established mathematical model of the floating photovoltaic power station's power collection system design to obtain the Pareto front, and then obtains the optimal design scheme for the floating photovoltaic power station's power collection system from the Pareto front based on the CWM-TOPSIS method, and generates the design drawing of the floating photovoltaic power station's power collection system.

[0150] To gain a further understanding and recognition of the structural features and achieved effects of the present invention, the following is a detailed description in conjunction with preferred embodiments and accompanying drawings:

[0151] Taking a floating PV power station in a certain water area as an example, the models and parameter settings of the relevant equipment of the floating PV power station are shown in Table 1, and the model of the collector cable is shown in Table 2.

[0152] Table 1 Models and Parameters of Relevant Equipment of Floating PV Power Station

[0153] Name Model Rated power of photovoltaic panel 550Wp Rated power of inverter 320kW Length of power generation unit 40m Width of power generation unit 5m Width of operation and maintenance passage 20m Fixed loss rate on DC side 88.7％ AC side conversion efficiency 94％ Annual peak hours 1548h Initial year attenuation rate of components 1％ Linear attenuation rate of components in the following year 3％ System operation years 25 years Cost of a single inverter 260,000 yuan Cost of a single photovoltaic panel 400 yuan / piece Power factor of photovoltaic panel 0.95

[0154] Table 2 Collector Cable Models

[0155]

[0156]

[0157] In China, each region divides light resources according to the annual equivalent utilization hours of PV. Among them, the annual equivalent utilization hours greater than 1600 hours are class I resource areas, those between 1400 - 1600 hours are class II resource areas, and those between 1200 - 1400 hours are class III resource areas. Considering the irradiation intensity in different regions, taking into account the land resources of the construction site and the effective utilization of PV modules, the upper limit requirements for the capacity ratio in different regions are formulated according to local conditions, and the selection range of the capacity ratio is determined, and there is:

[0158]

[0159] The annual equivalent utilization hours of the example water area PV is 1246h, belonging to class III resource area. Therefore, the selection range of its capacity ratio λ is [1 - 1.8].

[0160] (1) Comparison of Different Collector Cable Laying Schemes

[0161] First of all, different schemes were compared for the optimization of the collector cable laying of the floating PV power station. The specific schemes are as follows:

[0162] Set Case1 as the manual design scheme, Case2 as the optimized collector cable laying scheme integrating the improved FCM, DM - MSTP algorithms and the proposed automatic collector cable type selection algorithm, and Case3 as the collector cable laying scheme proposed by the present invention. The same on - road step - up substation and box - type substation coordinates are used for all three schemes. Among them, Case1 and Case2 use the collector cable models selected manually, as shown in Table 3; Case3 uses the collector cable models of the optimal design scheme generated based on the proposed scheme, as shown in Table 4.

[0163] Table 3 Types of collector cables selected manually

[0164] Model of collector line Current-carrying capacity (A) Price (yuan / m) Color of collector line cable ZC-YJHY23-26 / 3×120 245 180 Cyan ZC-YJHY23-26 / 3×185 310 213 Green ZC-YJHY23-26 / 3×240 360 235 Blue ZC-YJY23-26 / 3×300 520 800 Orange ZC-YJY23-26 / 3×400 590 1041 Red

[0165] Table 4 Types of collector cables used in Case 3

[0166] Model of collector line Current-carrying capacity (A) Price (yuan / m) Color of collector line cable ZC-YJHY23-26 / 3×120 245 180 Cyan ZC-YJHY23-26 / 3×240 360 235 Blue ZC-YJHY23-26 / 3×400 465 300 Yellow ZC-YJHY23-26 / 3×500 520 350 Purple

[0167] The laying diagrams of collector cables for different schemes are as follows Figure 3 to Figure 4 shown. Among them, Figure 3 is the schematic diagram of the collector cable laying scheme corresponding to Case 1; Figure 4 is the schematic diagram of the collector cable laying scheme corresponding to Case 2. Its box-type substation is divided into 6 areas, and the clustering center is represented by a five-pointed star, and Coordinate 1 is the location of the step-up substation on the road; Figure 5 is the schematic diagram of the collector cable laying scheme corresponding to Case 3. Its box-type substation is divided into 7 areas, and the clustering center is also represented by a five-pointed star, and Coordinate 1 is also the location of the step-up substation on the road. To more clearly and intuitively display the connection path of the collector cable, Figure 3 and Figure 4 adopt schematic diagrams and do not adopt the horizontal and vertical connection method. The connecting lines are in the corresponding connection order.

[0168] The comparison of the lengths and laying costs of collector cables for different schemes is shown in Table 5. Compared with Case 1, the reduction rates of the lengths of collector cables in Case 2 and Case 3 are 17.5% and 10.9% respectively, and the reduction rates of the investments in collector cables are 16.6% and 67.9% respectively, which proves that the scheme proposed by the present invention avoids the deficiencies of low automation degree and long design time in manual design, and can not only automatically select the optimal type of collector cable to realize the automatic laying of collector cables, but also effectively reduce the laying cost of collector cables in the floating photovoltaic power station.

[0169] Table 5 Comparison table of the lengths and laying costs of collector cables for different schemes

[0170] Case 1 Case 2 Case 3 Length of collector line cable (m) 24034 19823 21393 Laying cost of collector line cable (10,000 yuan) 1526.4 1273.2 488.5

[0171] (2) Comparison of different design schemes of the collector system in the floating photovoltaic power station

[0172] A comparative study was carried out on different design schemes of the collector system in the floating photovoltaic power station. Case 4 represents the manual design scheme, and Case 5 represents the single-objective sequential optimization scheme. The Case 5 scheme first optimizes the layout of the power generation units with the minimum layout cost of the power generation units, and then optimizes the laying of the collector cables with the minimum laying cost of the collector cables. Case 6 is the scheme proposed by the present invention. The types of collector cables for Case 4 and Case 5 are shown in Table 3, and the types of collector cables for Case 6 are shown in Table 6.

[0173] Table 6 Collector Cable Models for Case 6

[0174] Model of collector line Current-carrying capacity (A) Price (yuan / m) Color of collector line cable ZC-YJHY23-26 / 3×120 245 180 Cyan ZC-YJY23-26 / 3×275 275 320 Pink ZC-YJHY23-26 / 3×410 410 270 Black

[0175] The design result of the collector system for Case 4 is as follows Figure 6 shown. The capacitance ratio is 1.28. A total of 61 power generation units and 49 box-type substations are arranged in the water area.

[0176] The design result of the collector system for Case 5 is as follows Figure 7 shown. The capacitance ratio is 1.28. A total of 98 power generation units are arranged in the water area. Each power generation unit contains 5 minimum power generation units. The aspect ratio is 22:17. The box-type substations are divided into 6 areas in total. The clustering center is represented by a five-pointed star, and coordinate 1 represents the step-up substation on the road.

[0177] Figure 8 The blue dots in [Figure number] represent all feasible solutions for the multi-objective optimization of the collector system of the floating photovoltaic power station generated by artificial collaborative search and Pareto evaluation based on non-dominated sorting. Among all the feasible solutions, the Pareto front solution set is generated based on Pareto evaluation, as shown by the red dots. Then, the optimal solution is selected from the Pareto front solution set using CWM-TOPSIS, as shown by the green dots. Accordingly, the optimal design scheme for the collector system of the floating photovoltaic power station is the design scheme of the collector system of the floating photovoltaic power station corresponding to the green dot.

[0178] The design result of the collector system for Case 6 is as follows Figure 9 shown. The capacitance ratio is 1.3. A total of 128 power generation units are arranged in the water area. Each power generation unit contains 4 minimum power generation units. The aspect ratio is 36:33. The box-type substations are divided into 9 areas in total.

[0179] The comparison of the annual power generation, the layout cost of power generation units, and the laying cost of collector cables for different collector system design schemes is shown in Table 7. The power generation of Case 4 and Case 5 is calculated by Equation (9). Compared with Case 4 and Case 5, Case 6 realizes an increase in power generation of 14.4% and 11.7% respectively, and a decrease in the investment of collector cables of 41.8% and 32.8% respectively. It is confirmed that the scheme proposed in the present invention can not only automatically generate the multi-objective optimal design scheme of the collector system, avoid the problems of low automation in manual design and BIM design, but also automatically generate the optimal capacitance ratio and collector cable models, make fuller use of the water surface area, increase the annual power generation of the floating photovoltaic power station and reduce the laying cost of collector cables, thereby achieving the research goal of the multi-objective optimal design of the collector system of the floating photovoltaic power station.

[0180] Table 7 Comparison Table of Different Design Schemes of Collector System

[0181] Case4 Case5 Case6 Power generation (MW / year) 68303.4 69942.6 78123.5 Layout cost of power generation unit (10,000 yuan) 27244 27090.2 29934.3 Laying cost of collector line cable (10,000 yuan) 1526.4 1265.6 887.8

[0182] In summary, for the problem of multi-objective optimal design of the collector system of a floating photovoltaic power station, the present invention establishes a multi-objective optimal design model of the collector system of a floating photovoltaic power station with the layout cost of power generation units, annual power generation, and the laying cost of collector cables as the optimization objectives, while satisfying multiple constraint conditions. An integrated NSACS algorithm, Pareto evaluation, and a multi-objective optimal design scheme integrating CWM-TOPSIS are proposed. The scheme proposed by the present invention can not only generate the Pareto solution set frontier, but also automatically output the optimal design scheme of the collector system of the floating photovoltaic power station on the Pareto solution set frontier by means of CWM-TOPSIS. The present invention can effectively coordinate the layout cost of power generation units, annual power generation, and the laying cost of collector cables of the floating photovoltaic power station, so as to make full use of the water area space, and at the same time has the value of reference for actual engineering design and popularization and application.

[0183] The above embodiments are only used to describe the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A multi-objective optimization design method for the collector system of a floating photovoltaic power station, characterized in that The method includes the following steps: S1. Obtain the basic data of the floating PV power station; S2. Establish a multi-objective optimization design model for the collection system of the floating PV power station; S3. Conduct multi-objective optimization using the multi-objective optimization design model for the collection system of the floating PV power station to obtain a multi-objective optimization design scheme; S4. Based on the basic data of the floating PV power station and the multi-objective optimization design scheme, use the CWM-TOPSIS method to obtain the design result of the collection system of the floating PV power station.

2. The multi-objective optimization design method for the collection system of the floating PV power station according to claim 1, wherein in step S1, the basic data of the floating PV power station includes: light resource and water area data and device basic data; the light resource and water area data includes the light resource level, annual sunlight condition and water area condition of the location area of the floating PV power station; the device basic data includes the model and pricing of the inverter, the model and pricing of the PV panel, and the model of the collection cable.

3. The multi-objective optimization design method for the collection system of the floating PV power station according to claim 2, wherein step S2 specifically includes the following steps: S21. Based on the minimum power generation unit, establish a power generation unit layout cost model; S22. Establish an annual power generation model associated with the optimization variable capacity ratio of the collection system of the floating PV power station; S23. Based on the improved FCM clustering algorithm, DM-MSTP algorithm and the automatic selection and type selection method of the collection cable, establish a collection cable laying cost model; S24. Set the constraint conditions that need to be satisfied in the design of the collection system of the floating PV power station; S25. According to the power generation unit layout cost model, the annual power generation model, the collection cable laying cost model and the constraint conditions, establish a multi-objective optimization design model for the collection system of the floating PV power station.

4. The multi-objective optimization design method for the collection system of the floating PV power station according to claim 3, wherein step S3 specifically includes the following steps: S31. Combine the NSACS algorithm and the Pareto evaluation to generate the Pareto solution set frontier; S32. Based on the CWM-TOPSIS method, generate a multi-objective optimization design scheme of the optimal solution from the Pareto solution set frontier.

5. The multi-objective optimization design method for the collection system of the floating PV power station according to claim 4, wherein step S21 specifically includes the following steps: S211. Define a power generation unit composed of one inverter and n PV modules as the minimum power generation unit of the floating PV power station. Based on the definition of the minimum PV power generation unit, calculate the capacity of the minimum power generation unit and the number of PV modules in the minimum power generation unit using equations (1) and (2) respectively: P min = P INV × λ (1) In Equation (1), P min is the capacity of the minimum power generation unit, P INV is the rated power of the inverter, and λ is the capacity ratio; In formula (2), n is the number of photovoltaic modules in the minimum power generation unit; P min is the capacity of the minimum power generation unit; P gfb is the rated power of the photovoltaic panel; N zj is the number of photovoltaic panels in a photovoltaic module; n is an integer and is rounded using the rounding function; S212. Calculate the capacity of the power generation unit using equation (3); K a ×K b =n×N p.unit (3) In formula (3), the power generation unit is the basic component unit of the floating PV power station, which is composed of N p.unit minimum power generation units spliced together. Denote the arrangement of the PV modules in the power generation unit as a rows and b columns. Assume that there are K b PV modules in each row and K a PV modules in each column; N p.unit is the number of the minimum power generation units; S213. Calculate the length and width of the minimum power generation unit in the floating PV power station using equations (4) and (5). L = L unit × K b (4) W = W unit × K a (5) In Formula (4) and Formula (5), L and W respectively represent the length and width of the smallest power generation unit in the floating PV power station, L unit , W unit respectively represent the length and width of each PV module in the smallest power generation unit, and the length and width of each PV module are determined by the selected PV panel model, PV inclination angle, and the spacing between panels; K b represents the number of PV modules in each row in the power generation unit; K a represents the number of PV modules in each column in the power generation unit; S214. Set the capacity ratio of the floating PV power station, the number of minimum power generation units, and the aspect ratio of the power generation units as optimization variables, and set the layout cost of the power generation units of the floating PV power station as the objective function f1. Then, use Equation (6) to determine the objective function f1, and further determine the layout cost model of the power generation units according to the objective function f1: f1=(C INV +C gfb ×N zj ×n)×N p.unit ×N unit (6) Substitute Equation (1) and Equation (2) into Equation (6), then we have: In Equations (6) and (7), f1 is the objective function of the layout cost of the power generation unit, C INV is the cost of a single inverter, C gfb is the cost of a single photovoltaic panel, N zj is the number of photovoltaic panels in a photovoltaic module, n is the number of photovoltaic modules in the minimum power generation unit, N p.unit is the number of minimum power generation units, N unit is the number of power generation units of the floating photovoltaic power station, P INV is the rated power of the inverter, λ is the capacity ratio, P gfb is the rated power of the photovoltaic panel.

6. The multi-objective optimization design method for the collector system of the floating PV power station according to claim 5, characterized in that: The step S22 specifically includes the following steps: S221. Use Equation (8) to calculate the total photovoltaic power generation of the floating PV power station: In formula (8), P out is the total photovoltaic power generation within the unit time stamp; PR is the energy efficiency ratio; P0 is the standard installed capacity, that is, the installed capacity under the 1:1 capacity ratio; G i,k is the cumulative irradiation total on the inclined plane within the unit time stamp, with the unit of kWh / m 2 ; τ k is the unit time stamp; G i,ref is the standard irradiance; S222. Decouple the power generation formula (8), establish the relationship between the power generation and the capacity ratio, and set the negative value of the annual power generation of the floating PV power station as the objective function f2. Based on the objective function f2, establish an annual power generation model associated with the capacity ratio of the optimization variable of the collector system of the floating PV power station: f2 = -P out = -PR0 × PR1 × T ref × P o × λ × [1 - a - b × (y - 1)] (9) P0 = P INV × N p.unit × N unit (10) In formulas (9) and (10), f2 is the objective function of the annual power generation of the floating PV power station; P out is the total PV power generation within a unit time stamp; PR0 is the fixed loss rate on the DC side, representing the DC side losses other than the light rejection rate and the module attenuation rate; PR1 is the AC side conversion efficiency; T ref is the local annual peak hours, which can be queried through the historical meteorological database; P0 is the standard installed capacity, that is, the installed capacity under a 1:1 capacity ratio; λ is the capacity ratio; a is the first-year attenuation rate of the module; b is the second-year linear attenuation rate of the module; y is the number of years of system operation; P INV is the rated power of the inverter; N p.unit is the number of the minimum power generation units; N unit is the number of power generation units of the floating PV power station.

7. The multi-objective optimization design method for the collector system of the floating PV power station according to claim 6, characterized in that: The S23 specifically includes the following steps: S231. Set the collector cable type as the optimization variable and use Equation (11) to establish the constraint condition: In formula (11), S q·f is the capacity of the f-th box-type substation towed by the q-th collector cable, S q·max represents the maximum apparent power of the q-th collector cable, F represents the number of box-type substations in the q-th collector cable, Q is the number of collector cables in the collector system, I represents the current-carrying capacity of the collector cable, k represents the number of towed box-type substations, P N is the rated power of the power generation unit of the floating PV power station, U is the voltage of the collector system, cosθ is the power factor of the PV panel, N max is the maximum number of box-type substations that a single collector cable can withstand, I max is the maximum current-carrying capacity of the collector cable; the constraint conditions include the capacity constraint of the collector cable, the current-carrying capacity constraint, and the constraint on the maximum number of towable box-type substations; P N represents the rated power of the power generation unit of the floating PV power station; S232. Take the onshore substation as the radiation center, partition the box-type substations by using the improved FCM clustering algorithm, and based on the constraint condition, use Equation (12) to calculate the distances between the box-type substations and between the box-type substations and the onshore substation in each partition, and construct a distance matrix; In formula (12), L d is an element in the distance matrix of the d-th partition; are the horizontal and vertical coordinates of the box-type substation in the d-th partition respectively, where d = 1, 2, …, N, i = 1, 2, …, m; S233. Based on the constraint condition and the distance matrix, use the DM-MSTP algorithm to optimize the length of the collector cables of the floating PV power station, then use the automatic cable type selection algorithm for collector cables to optimize the cable type selection of the collector cables, determine the laying scheme of the collector cables of the floating PV power station, and finally establish a laying cost model of the collector cables based on the laying scheme of the collector cables of the floating PV power station; The S233 specifically includes the following steps: S2331. Based on the constraint condition and the distance matrix, use the DM-MSTP algorithm to optimize the length of the integrated cables of the floating PV power station, and generate a collector cable connection table for each partition; the first row of the inherited cable connection table represents the starting node, the second row represents the ending node, and the third row represents the distance between the starting node and the ending node; S2332. Based on the integrated cable connection table, calculate the number of times each node appears. If the number of times each node appears is less than or equal to 2, record the number of box-type substations mounted on each node, and perform collector cable type selection according to the constraint condition of Equation (11), and execute step S2335; if there is a node among all nodes whose number of appearances is greater than 2, execute step S2333; S2333. Find the node with an occurrence count of 1 in the collector cable connection table except for the substation on the line. Denote the position of this node in the table as s, and let r = s - 1. Compare the values of Link(1, s) and Link(2, r), where Link(1, s) represents the element in the first row and the s-th column of the table, and Link(2, r) represents the element in the second row and the r-th column of the table. If Link(1, s) = Link(2, r), then record the number of box-type substations mounted on the node Link(2, r). Then, decrement the value of s by 1 and the value of r by 1, and continue to record the corresponding number of box-type substations mounted. If Link(1, s) ≠ Link(2, r), then decrement the value of r by 1, and continue to determine whether Link(1, s) is equal to Link(2, r - 1). Traverse forward in this way until the serial number is 1, and then end the traversal process. Record all the above connection processes in the same cell array, and denote this array as LJ. After completion, proceed to step S2334; S2334. In the collector cable connection table, find the nodes with an occurrence count greater than 2, and record the found nodes as u. For the cell array LJ, merge the arrays that contain u. The merging rule is to first determine the position of u in the cell array, and then compare the number of box-type substations mounted on the u node in different arrays that contain u. Add the number of box-type substations mounted on the u node with the smaller number of box-type substations mounted to the u node with the larger number of box-type substations mounted. Continuously merge according to this rule until there is only one cell left in the cell array LJ. At this time, according to the constraint conditions specified in Equation (11), use the automatic selection algorithm for collector cables to optimize the selection of collector cables; S2335. According to the selection of collector cables, that is, the selection of chain-shaped and tree-shaped collector cables, determine the laying scheme of the collector cables for the floating photovoltaic power station, and then establish a collector cable laying cost model using Equation (13). In Equation (13), f3 is the objective function of the cost of laying collector cables, N is the number of partitions of fuzzy C-clustering, m is the number of box-type substations in each partition, L C.q is the length of the q-th collector cable, C C.q (φ) is the cost per kilometer of the q-th collector cable, and φ is the cross-sectional area of the collector cable.

8. The multi-objective optimization design method for the collector system of a floating photovoltaic power station according to claim 7, characterized in that In step S25, the multi-objective optimization design model of the collector system of the floating photovoltaic power station sets the capacity ratio, the number of minimum power generation units, the aspect ratio of the power generation unit length and width, and the collector cable model as optimization variables, and sets the layout cost of the power generation units of the floating photovoltaic power station, the annual power generation, and the collector cable laying cost as objective functions; the constraint conditions of the multi-objective optimization design model of the collector system of the floating photovoltaic power station include: all power generation units shall not exceed the water area boundary, there shall be no intersection and overlap between power generation units, the width reserved for the operation and maintenance passage meets the national standard requirements, and at the same time, the collector cable meets its capacity limit condition, that is, the maximum carrying capacity of the selected collector cable model must be greater than or equal to the sum of the capacities of the box-type substations it drives downstream; The multi-objective optimization design model of the collector system of the floating photovoltaic power station is shown in formula (14): In Equation (14), x i,v , y i,v are the horizontal and vertical coordinates of each vertex of the power generation unit, where i = 1, 2... m represents the serial number of the power generation unit, and v = 1, 2, 3, 4 represents the serial number of the vertex of the power generation unit; x min , x max , y min , y max represent the boundary coordinates of the water area respectively, L space is the width of the operation and maintenance passage, S q·f is the capacity of the f-th box-type substation towed by the q-th collector cable, S q·max represents the maximum apparent power of the q-th collector cable, F represents the number of box-type substations in the q-th collector cable, Q is the number of collector cables in the collector system, I represents the current-carrying capacity of the collector cable, k represents the number of towed box-type substations, P N is the rated power of the power generation unit of the floating PV power station, U is the voltage of the collector system, cosθ is the power factor of the PV panel, N max is the maximum number of box-type substations that a single collector cable can withstand, I max is the maximum current-carrying capacity of the collector cable.

9. The multi-objective optimization design method for the collector system of a floating photovoltaic power station according to claim 8, characterized in that Step S31 specifically includes the following steps: S311. Initialize the population and conduct fitness evaluation S3111. Create initial populations for two super individuals A and B respectively using Equation (15), and calculate the fitness of each individual of super individuals A and B (f1(x), f2(x)....f n (x)); S3112. Sort each individual of A and B using the fitness function value, and calculate the crowding distance D of each individual using Equation (16). c Then store the fitness, ranking, and CD of each individual in the A and B super individuals; In Equations (15) and (16), p = 1... M, q = 1... D, where M is the population size, D is the problem dimension, rand is a random number uniformly distributed, low q and up q are the search lower and upper bounds of q respectively, y p,A and y p,B are the fitness values generated by two super individuals A and B respectively, fitnessA and fitnessB are the fitness values of A and B respectively, D c (A1) is the crowding degree of individual A1, A2 and A3 are two individuals adjacent to A1, f l (A2) and f l (A3) are the objective function values of individuals A2 and A3 respectively, A0 is the neighborhood center of individuals A2 and A3, f l (A0) is the objective function value of the neighborhood center; S312. Selection of predators and prey For the super individuals A and B in each parallel computing area, randomly select predators and prey using Equation (17): In Equation (17), a and b are randomly generated numbers, predator represents the predator, prey represents the prey, := represents the update operation, and y predator and y prey and y A and y B represent the fitness values generated for the corresponding predator, prey, A, and B, respectively; S313. Generation of offspring S3131. Utilize the biological interaction between the predator and the prey to generate the new position and direction of the predator, and simultaneously implement boundary control; S3132. Select and crossover between predators to enhance the hunting ability. Based on the hunting ability, determine the cooperation probability between predators using Equation (18); S3133. Dynamically adjust the dynamic factors for searching positions using Equation (19) to generate predator offspring. The dynamic factors include the mutation rate and selection crossover rate of the search position; In Equations (18) and (19), x is the mutation matrix, predator represents the predator, prey represents the prey, a, b, and c are randomly generated numbers, P is the control parameter, x p,q 、stronger p,q 、predator p,q are the mutation matrix value, stronger predator, and predator at the corresponding (p, q);. represents the dot product operation of matrices or vectors; S3134. Calculate and store offspring information. Calculate the fitness, rank, and crowding distance CD of the offspring, and store this information in a separate location; S314. Elite selection Find the optimal population by combining predators and offspring, sort the combined population according to fitness and rank, and select the optimal individuals; S315. Repeat steps S311 - S314 until the maximum number of iterations to generate the Pareto solution set frontier for the multi-objective optimization of the collector system of the floating PV power station; The specific steps of step S32 are as follows: S321. Based on the subjective weight and the objective weight Use Equation (20) to obtain the combined weight ω * : In formula (20), ω * is the combined weight, α1 and α2 are the weight coefficients, is the subjective weight, is the objective weight; S322. Use the concept of game theory shown in Equation (21) to find the Nash equilibrium point: Min||ω * -ω k ||2, k = 1, 2 (21) In formula (21), ω * is the combined weight, and ω k represents the transpose of the weight, where k = 1, 2; S323. According to the properties of matrix differentiation, obtain that the optimal solution of Equation (21) satisfies the conditions shown in Equation (22): In Equation (22), is the subjective weight, is the objective weight, α1 and α2 are weight coefficients, and ω1 and ω2 are the transposes of the subjective weight and the objective weight; S324. Obtained through normalization processing And the optimized combined weight is calculated using Equation (23): In formula (23), ω * is the combined weight, and are the normalized weight coefficients, is the subjective weight, is the objective weight; S325. Use TOPSIS to evaluate the generated Pareto frontier. After obtaining the optimized combined weights, use TOPSIS to evaluate the generated Pareto solution set frontier; In S325, the evaluation of the generated Pareto solution set frontier using TOPSIS includes: S3251. Assume there are m objects and n evaluation criteria, and construct a decision matrix; S3252. Normalize and weight the decision matrix to obtain the weighted decision matrix V shown in Equation (24): In Equation (24), V is the weighted decision matrix, and b ij is the corresponding weight, where i = 1...m, j = 1...n; S3253. Use the weighted decision matrix V to evaluate the generated Pareto solution set frontier; S326. Determine the positive ideal solution $F^+$ using equations (25) and (26) + and the negative ideal solution $F^-$ - , and calculate the best distance and the worst distance of each solution using equations (27) and (28): F + = max 1≤i≤m v ij (25) F - = min 1≤i≤m v ij (26) In formulas (25) to (28), F + and F - are the positive ideal solution and the negative ideal solution respectively, v ij is the value at (i, j) in the weighted decision matrix, and are the best distance and the worst distance respectively, F i + and F i - are the positive ideal solution and the negative ideal solution of the i-th row in the weighted decision matrix V respectively, F ij is the ideal solution value at (i, j); S327. Calculate the comprehensive evaluation index C using Equation (29) j , and based on the comprehensive evaluation index C j sort the solutions to determine the optimal solution of the Pareto front, and then determine the optimal design scheme of the collector system of the floating PV power station: In formula (29), C j is the comprehensive evaluation index, and are the best distance and the worst distance respectively.

10. The multi-objective optimization design method for the collector system of the floating PV power station according to claim 9, characterized in that The specific steps of S4 are as follows: S41. Based on the available water area, water restricted areas, and operation and maintenance passage requirements, obtain the layable areas of the power generation units of the floating PV power station, and based on national standards requirements and design resources, obtain the PV panel models, inverter models, collector cable groups, and number of connection components; S42. Initialize the capacity ratio λ, the minimum number N of superposed power generation units, p.unit the aspect ratio K of the power generation unit, and the type of collector cable, set the number of initial super individuals to 200, and the number of iterations to 30; S43. According to the capacity ratio λ and the minimum number of power generation units N in the super individual p.unit , the aspect ratio K = L / W value of the water surface photovoltaic power generation unit and the constraint conditions in formula (13), lay the power generation units in the water area space: If the group of λ, N p.unit , and K = L / W value in the super individual cannot achieve the laying of the power generation units, directly jump to the next group of super individuals; if the laying of the power generation units can be achieved, calculate the objective function f1 using formula (7), calculate the objective function f2 using formula (9), and generate the coordinates of the box-type substation according to the setting that the box-type substation is arranged at the center of the power generation unit; according to the type of the collector cable in this group of super individuals, based on the constraint conditions in formula (13), use the improved FCM algorithm to partition the box-type substation with the substation on the road as the base point, and then integrate the DM-MSTP algorithm and the proposed automatic selection algorithm for the collector cable to lay the collector cable; if the collector cable in this group of super individuals is not sufficient to tow the box-type substation in any partition, jump to the next group of super individuals; if it can tow the box-type substation in all partitions, calculate the objective function f3 according to formula (12); S44. Terminate the iteration when the number of iterations is reached to generate the Pareto frontier for the multi-objective optimization of the collector system of the floating PV power station; S45. Based on the CWM-TOPSIS method, output the optimal design scheme of the collector system of the floating PV power station from the generated Pareto solution set frontier for the multi-objective optimization of the collector system of the floating PV power station, and obtain the design results of the collector system of the floating PV power station.

Images