Wind-light-water storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty

Abstract

The invention discloses a wind-light-water storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty, and the method comprises the steps: building a wind-light-water storage multi-energy system complementary coordinated optimization scheduling model, and constructing constraint conditions for the wind-light-water storage multi-energy system complementary coordinated optimization scheduling model; wherein in the wind, light and water storage multi-energy system complementary coordination optimization scheduling model, the optimization targets with the minimum net load fluctuation, the maximum energy storage system operation income and the minimum cost belong to the upper-layer model, and the optimization targets with the minimum thermal power generating unit operation cost and the maximum renewable energy consumption capability belong to the lower-layer model; power balance constraints in constraint conditions are rewritten into strict power equation balance constraints, wind power fuzzy parameters, photovoltaic fuzzy parameters and load fuzzy parameters are introduced at the same time, and the strict power balance constraints are relaxed into power constraint balance under a predetermined confidence level condition; the possibility that the balance constraint condition is established is not less than a preset confidence level, so that an uncertain factor set is constructed; a layered optimization scheduling scheme is adopted for the wind-light-water storage multi-energy system complementary coordination optimization scheduling model with the introduced fuzzy parameters, an upper layer model is solved through a solver, and a lower layer model is solved through an optimization algorithm. According to the method, uncertainty modeling, hierarchical collaborative optimization, multi-subject benefit coordination and an advanced intelligent algorithm are fused, so that the overall peak regulation capacity and the renewable energy consumption level of the wind-light-water storage system are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to a wind-solar-water storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty, and belongs to the field of new energy power system dispatching optimization. BACKGROUND

[0002] In recent years, with the rapid development of the electric power industry, the proportion of renewable energy such as wind power and photovoltaic power is increasing. The anti-peak regulation and uncertain characteristics of wind power have widened the peak-valley difference of load, and the double pressure of source and load has increased the peak regulation burden of the power system. In order to cope with the phenomenon that the fluctuation of wind power output will lead to frequent start-stop of thermal power units, for example, when wind power is generated in large quantities, measures such as shutting down efficient thermal power units may be taken to fully accommodate wind power; if the wind power is limited to be connected to the grid, a large amount of wind power will be abandoned, which wastes valuable renewable resources. It is difficult to meet the demand for renewable energy consumption and power system peak regulation by relying solely on the existing regulation capacity of the system, so it is necessary to establish a multi-source complementary coordination mechanism to fully tap the flexible regulation capacity of the power system, meet the demand for peak regulation, and improve the consumption of renewable energy. At the same time, considering the influence of natural conditions and other unknown factors, there are still errors between the predicted results and the actual power output and load power.

[0003] Current researches generally adopt diversified modeling methods, from early deterministic optimization models based on linear or nonlinear programming, to the introduction of robust optimization, stochastic programming and fuzzy optimization to cope with the strong uncertainty of wind power and photovoltaic output; at the same time, intelligent methods such as deep reinforcement learning (such as TD3, DDPG) and meta-heuristic algorithms (such as genetic algorithm, particle swarm optimization) are also widely used to solve high-dimensional, non-convex, nonlinear complex scheduling problems. In terms of optimization objectives, multi-objective coordination has become the mainstream trend, typical objectives include minimizing total system operating cost, maximizing renewable energy consumption, reducing net load fluctuation, improving energy storage economic benefits and ensuring water and electricity ecological constraints, etc., which are often coordinated through weighted summation and Pareto frontier analysis. At the same time, the modeling of internal coupling mechanisms is increasingly refined, such as considering the influence of water turbine efficiency curve, reservoir dynamic storage and water head change on water power regulation capacity, depicting the deep peak shaving characteristics and start-stop flexibility of thermal power units, and integrating different forms of energy storage such as electrochemical energy storage and pumped storage into a unified scheduling framework. In addition, research gradually extends from standard test systems (such as IEEE30 node) to actual clean energy bases (such as the upper reaches of the Yellow River and the Yalong River basin), promoting the transformation of theoretical achievements to engineering applications. However, there are still some key deficiencies in current research: in multi-objective optimization, different physical dimensionless objective functions are often subjectively weighted, lacking scientific normalization or objective weighting mechanism, resulting in poor robustness and weak comparability of the results; the modeling of wind and light output uncertainty is still rough; in terms of algorithm, traditional mathematical programming methods are prone to "dimension disaster", while data-driven reinforcement learning has strong generalization ability, but it faces the problems of strong training dependence and poor strategy explainability, which is difficult to meet the strict requirements of power system safety and authentication.

[0004] In view of the problems and challenges existing in the wind, light, water and storage multi-energy complementary system optimization scheduling and decision-making model, future research needs to build a multi-objective optimization scheduling model that considers the influence of natural conditions and other unknown factors, so that the prediction and actual power output and load power error are too large, and the deep peak shaving capacity of thermal power units is fully utilized, and the peak clipping and valley filling capacity of energy storage devices is utilized to improve the flexibility of the system and promote the consumption of renewable energy. SUMMARY

[0005] In view of the problems of the system affected by natural and other uncertain factors after large-scale renewable energy such as wind power and photovoltaic is connected to the grid, the present application provides a wind, light, water and storage multi-energy system complementary peak regulation and scheduling method considering source and load uncertainty, to establish a wind, light, water and storage multi-energy system complementary coordination optimization scheduling model introducing fuzzy parameters; further, a traditional solver is used to solve the model combined with an optimization algorithm based on an improved multi-objective pelican algorithm, to achieve the purpose of improving the renewable energy consumption level and the economic efficiency of thermal power units.

[0006] The technical scheme of the present application is:

[0007] According to a first aspect of the present application, a wind-solar-hydro-storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty is provided, comprising:

[0008] Step 1, a wind-solar-hydro-storage multi-energy system complementary coordinated optimization scheduling model is established, with the minimum net load fluctuation, the maximum storage system operation benefit and the minimum cost, the minimum operation cost of thermal power units and the maximum renewable energy consumption capacity as the objective function, and constraint conditions are constructed for the wind-solar-hydro-storage multi-energy system complementary coordinated optimization scheduling model; wherein the optimization objectives of the minimum net load fluctuation, the maximum storage system operation benefit and the minimum cost belong to the upper model, and the optimization objectives of the minimum operation cost of thermal power units and the maximum renewable energy consumption capacity belong to the lower model; the constraint conditions include power balance constraint, thermal power unit constraint, wind power output constraint, hydropower station output constraint, photovoltaic power station output constraint, storage constraint and line transmission capacity constraint;

[0009] Step 2, the power balance constraint in the constraint condition is rewritten as a strict power equation balance constraint, while introducing wind power fuzzy parameters, photovoltaic fuzzy parameters and load fuzzy parameters, the strict power balance constraint is relaxed to a power constraint balance under a predetermined confidence level, so that the possibility of the balance constraint condition being true is not less than the predetermined confidence level, thereby constructing an uncertain factor set;

[0010] Step 3, the wind-solar-hydro-storage multi-energy system complementary coordinated optimization scheduling model with fuzzy parameters is introduced, and a hierarchical optimization scheduling scheme is adopted, the upper model is solved by a solver, and the lower model is solved by an optimization algorithm.

[0011] Further, the net load fluctuation minimum objective function expression is as follows:

[0012] ;

[0013] In the formula: is a scheduling period; is the net load value at time t; is the average value of the net load in a scheduling period.

[0014] Further, the storage system operation benefit maximum and cost minimum objective function expression is as follows:

[0015] ;

[0016] In the formula: is the storage system operation benefit; is the storage system operation cost.

[0017] Further, the expression of the minimum operation cost objective function of the thermal power unit is:

[0018] ;

[0019] In the formula: is the operation cost of the conventional peak regulation of the thermal power unit; is the total number of the thermal power units; is the total number of the thermal power units is the loss cost; is the oil injection cost of the t-th thermal power unit at the t-th time; is the oil injection cost of the t-th thermal power unit at the t-th time; is a scheduling period; and are the minimum technical output and the maximum output of the thermal power unit, respectively; is the stable combustion limit load value of the oil injection depth peak regulation of the thermal power unit; is the stable combustion load value of the non-oil injection depth peak regulation of the thermal power unit; is the output of the t-th thermal power unit at the t-th time; is the output of the t-th thermal power unit at the t-th time;

[0020] Further, the expression of the maximum renewable energy consumption capacity is:

[0021] ;

[0022] In the formula: is a scheduling period; is the total number of the wind power plants; is the total number of the hydropower stations; is the total number of the photovoltaic power stations; represents the wind power plant that discards the wind power at the t-th time; is the photovoltaic power station that discards the light power at the t-th time; is the hydropower station that discards the water power at the t-th time; is the time length of each time period.

[0023] Further, the upper model is solved by the solver Cplex, and the lower model is solved by the multi-objective pelican optimization algorithm.

[0024] According to the second aspect of the present application, a wind-solar-hydro-storage multi-energy system complementary peak regulation scheduling system considering source-load uncertainty is provided, which comprises the module of the method in any one of the above.

[0025] According to the third aspect of the present application, a terminal device is provided, which comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, and the processor is configured to execute the steps of the method in any one of the above.

[0026] The beneficial effects of the present application are: the present application, by fusing uncertainty modeling, hierarchical collaborative optimization, multi-agent benefit coordination and advanced intelligent algorithm, not only effectively improves the overall peak regulation capability and renewable energy consumption level of the wind-solar-hydro-storage system, but also significantly reduces the system operation cost and the adjustment burden of the thermal power unit, enhances the operation flexibility, economy and reliability of the power system under the dual uncertainty of source and load, and provides a technical path with both theoretical value and engineering practicality for constructing a high proportion of renewable energy power system. BRIEF DESCRIPTION OF DRAWINGS

[0027] Fig. 1 is a flow chart of the present application.

[0028] Fig. 2 is a flow chart of the multi-objective pelican optimization algorithm of the present application. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be described clearly and completely below with reference to the drawings. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other in any manner without conflict.

[0030] Embodiment 1: As shown, according to the first aspect of the embodiments of the present application, a wind-solar-hydro-storage multi-energy system complementary peak regulation scheduling method considering source and load uncertainty is provided, comprising: Figs. 1-2 Step 1, considering that the multi-energy system has complex power supply structure and involves many variables and constraint conditions, a hierarchical optimization scheduling scheme is adopted, a wind-solar-hydro-storage multi-energy system complementary coordination optimization scheduling model with the minimum net load fluctuation, the maximum operating income and the minimum cost of the energy storage system, the minimum operating cost of the thermal power unit and the maximum renewable energy consumption capacity of wind and solar as the objective functions is established, and constraint conditions are constructed for the wind-solar-hydro-storage multi-energy system complementary coordination optimization scheduling model; wherein the optimization objectives of the minimum net load fluctuation, the maximum operating income and the minimum cost of the energy storage system belong to the upper model, and the optimization objectives of the minimum operating cost of the thermal power unit and the maximum renewable energy consumption capacity of wind and solar belong to the lower model; the constraint conditions include power balance constraint, thermal power unit constraint, wind power output constraint, hydropower station output constraint, photovoltaic power station output constraint, energy storage constraint and line transmission capacity constraint.

[0031]

[0032] ​Step 2, since the system power balance contains uncertain variables, the power balance under the determined condition will no longer be applicable, therefore, when arranging the system day-ahead scheduling output, the influence of the system uncertainty factors should be considered. A fuzzy processing method is adopted, the power balance constraint in the constraint condition is rewritten as a strict power equation balance constraint, and wind power fuzzy parameters , photovoltaic fuzzy parameters , load fuzzy parameters , the strict power balance constraint is relaxed to the power constraint balance under the predetermined confidence level , the possibility that the balance constraint condition is established is not less than , so as to construct the uncertain factor set;

[0033] Step 3, the hierarchical optimization scheduling scheme is adopted for the wind-solar-hydro-storage multi-energy system complementary coordination optimization scheduling model with the introduction of fuzzy parameters, the upper model is solved by the solver Cplex, and the lower model is solved by the multi-objective pelican optimization algorithm. The scheduling strategy solved by the lower model is judged: whether the profits of each thermal power unit and the profits of the wind power unit are greater than or equal to 0, and when both of them are greater than 0, the scheduling strategy is taken as the final scheduling strategy; otherwise, the thermal power unit is adjusted to reduce the peak shaving space and then re-solved.

[0034] Further, the Levitation Flight mechanism is added to the multi-objective pelican optimization algorithm, and the improved multi-objective pelican algorithm is used to solve the global optimal solution.

[0035] Further, the step 1 specifically comprises:

[0036] The present application is based on a wind-solar-hydro-storage combined system, and the minimum net load fluctuation, the maximum operating benefit and the minimum cost of the energy storage system, the minimum operating cost of the thermal power unit and the minimum renewable energy curtailment are taken as optimization targets respectively, and a wind-solar-hydro-storage multi-energy system complementary coordination optimization scheduling model is established.

[0037] 1, objective function.

[0038] 1) minimum net load fluctuation objective function.

[0039] The net load is the actual remaining load of the thermal power unit after deducting the wind-solar-hydro-storage combined output in the system. In order to fully utilize the capacity of the energy storage system to compensate the fluctuation of the wind-solar power output, the fluctuation of the net load of the thermal power unit is minimized as much as possible, and the thermal power unit is prevented from adjusting the output frequently, and the following minimum net load fluctuation objective function is established:

[0040] (10)

[0041] (11)

[0042] (12)

[0043] In the formula: is a scheduling period; is the net load value at time t; is the average value of the net load in a scheduling period; is the total load value at time t; is the actual grid-connected power of the wind farm at time t; is the actual grid-connected power of the wind farm at time t; is the actual grid-connected power of the hydropower at time t; is the actual grid-connected power of the hydropower at time t; is the actual grid-connected power of the photovoltaic at time t; is the actual grid-connected power of the photovoltaic at time t; is the discharge power of the energy storage device at time t; represents the discharge of the energy storage device; represents the charging of the energy storage device; is the total number of wind farms; is the total number of photovoltaic power stations; is the total number of hydropower stations.

[0044] 2) The objective function of maximizing the operating income and minimizing the cost of the energy storage system.

[0045] 2.1) Operating income model of the energy storage system.

[0046] During the operation of the energy storage system, the operating electricity income and environmental income of the energy storage system are the main factors affecting its operation economy. The operating electricity income of the energy storage system is mainly considered to establish the operating income and environmental income model of the energy storage system.

[0047] (13)

[0048] In the formula: is the operating income of the energy storage system; is the electricity price of the grid; and respectively represent the charging and discharging efficiency of the energy storage system; and are the charging and discharging power of the energy storage system at time t; M is the total number of pollutant types produced by the upper grid; is the unit emission cost of the pollutant; is the density of the kth pollutant emitted per unit of electricity produced by the upper grid.

[0049] 2.2) Operating cost model of the energy storage system.

[0050] The operating cost model of the energy storage system established in this invention mainly considers the charging and discharging costs of the energy storage system.

[0051] (14)

[0052] In the formula: For the operating costs of energy storage systems; The charging and discharging power cost coefficient of the energy storage system; and These are the charging and discharging power of the energy storage system at time t, respectively.

[0053] 2.3) Based on the energy storage system operation revenue model and energy storage system operation cost model, the objective function for maximizing energy storage system operation revenue and minimizing cost is constructed as follows:

[0054] (15)

[0055] 3) Objective function for minimizing the operating cost of thermal power units.

[0056] 3.1) Cost model for conventional peak-shaving operation of thermal power units.

[0057] (16)

[0058] In the formula: Costs of routine peak-shaving operation of thermal power units; and These are the coal consumption cost and start-up / shutdown cost of thermal power units, respectively. , and thermal power units Consumption coefficient; For time t, the first The output of the thermal power unit; This represents the total number of thermal power units. For thermal power units at time t Peak-shaving operating costs; For thermal power units at time t The running status;

[0059] 3.2) Peak-shaving unit loss cost model.

[0060] When the generating unit is undergoing peak shaving, its operating state deviates significantly from the design value, resulting in a substantial decrease in power generation efficiency and incurring additional unit loss costs and oil injection costs. This invention roughly calculates the unit loss costs using the most commonly used Manson-Coffin formula, as follows:

[0061] (17)

[0062] In the formula: For thermal power units Loss cost, β is the operation influence coefficient of thermal power units; For the purchased cost of the nth thermal power unit; Nfrepresents the rotor cracking cycle number determined by the rotor low-cycle fatigue curve.

[0063] 3.3) Oil injection cost for stable combustion of the peak-regulating unit.

[0064] In the oil injection peak-regulating stage, the additional oil injection peak-regulating cost generated by the oil injection combustion support is as follows:

[0065] (18)

[0066] In the formula: Ctis the oil injection cost of the nth thermal power unit at time t, Ctis the oil injection cost of the nth thermal power unit at time t, Qtis the oil injection amount of the nth thermal power unit at time t in the peak-regulating stage; Qtis the oil injection amount of the nth thermal power unit at time t in the peak-regulating stage; Qtis the oil injection amount of the nth thermal power unit at time t in the peak-regulating stage;

[0067] In summary, the operation cost of the thermal power unit is segmented according to different operation states as follows:

[0068] (19)

[0069] In the formula: and are the minimum technical output and the maximum output of the thermal power unit, respectively; Ctis the stable combustion limit load value of the unit in the deep peak-regulating stage with oil injection; Ctis the stable combustion load value of the unit in the deep peak-regulating stage without oil injection.

[0070] 4) Renewable energy consumption capacity

[0071] The renewable energy consumption capacity is represented by the sum of the abandoned wind power, the abandoned water power, and the abandoned light power in the dispatching period. The more the abandoned wind power, the abandoned light power, and the abandoned water power, the weaker the renewable energy consumption capacity.

[0072] (20)

[0073] In the formula: Pwind,trepresents the abandoned wind power of the wind farm at time t; Ppv,trepresents the abandoned light power of the photovoltaic power station at time t; Pwater,trepresents the abandoned water power of the hydropower station at time t; is the length of each time period, which is in this article.

[0074] 2. Constraints.

[0075] 1) Power balance constraint.

[0076] (21)

[0077] where, Pi(t) is the active power output of the ith thermal power unit at time t; Pwi(t) is the actual grid-connected power of the wind farm i at time t; Pwi(t) is the actual grid-connected power of the wind farm i at time t; Ph(t) is the actual grid-connected power of the hydro power unit h at time t; Ppv(t) is the actual grid-connected power of the photovoltaic m at time t; Pst(t) is the discharging power of the energy storage device t at time t, Pst(t) represents the discharging of the energy storage device, Pst(t) represents the charging of the energy storage device; Pd(t) is the load of node j at time t; N is the total number of load nodes.

[0078] 2) Thermal power unit constraints.

[0079] Conventional peak regulation unit:

[0080] (22)

[0081] Deep peak regulation unit:

[0082] (23)

[0083] Unit ramping constraint:

[0084] (24)

[0085] where, and are the maximum upward ramping rate and the maximum downward ramping rate of the thermal power unit, respectively.

[0086] 3) Wind power output constraint.

[0087] (25)

[0088] where, Pwi(t) is the maximum output of the wind farm i at time t.

[0089] 4) Hydroelectric power station output constraint.

[0090] (26)

[0091] where, ​is the maximum output of hydroelectric generator h at time t.

[0092] 5) Photovoltaic power plant output constraint.

[0093] (27)

[0094] where, is the maximum output of photovoltaic power plant m at time t.

[0095] 6) Energy storage constraint.

[0096] Energy storage constraint:

[0097] (28)

[0098] where: is the state of charge of energy storage system at time t; represents the self-discharge rate of energy storage; and are the charging and discharging power of energy storage system at time t, respectively; and represent the charging and discharging efficiency of energy storage system, respectively; is the capacity of energy storage system; and are the upper and lower limits of state of charge of energy storage system, respectively.

[0099] Energy storage charging and discharging constraint:

[0100] (29)

[0101] where: and are the charging and discharging state of energy storage system, respectively; and are the minimum and maximum values of charging power, respectively; and are the minimum and maximum values of discharging power, respectively.

[0102] 7) Line transmission capacity constraint.

[0103] The DC power flow is used to express the network security constraint:

[0104] (30)

[0105] where: is the transfer distribution factor of node M to line L. is the active power injection of node M to scenario s at time t; U is the number of nodes; is the maximum active power injection of line L; is the active power injection of line L to scenario s at time t.

[0106] Further, the step 2 specifically comprises:

[0107] To ensure the safe and stable operation of the power system, the load balance within the system must be maintained, and sufficient rotating reserve capacity must be configured. Therefore, when formulating the day-ahead scheduling output plan of each generator, the equality constraints (such as power balance) and inequality constraints (such as the upper and lower limits of unit output, ramp rate, etc.) of the system must be strictly met. Although various methods and technologies have been developed for power prediction of uncertain power sources such as wind power, photovoltaic power and load, due to the influence of weather conditions and other uncontrollable external factors, there is still a certain deviation between the prediction results and the actual output or load. Especially in the scenario of large-scale wind power and photovoltaic power grid connection or high load level, such prediction errors cannot be ignored and may have a significant impact on the safe operation of the system. In the day-ahead scheduling model of the power system, the wind power output, photovoltaic power output and load power are all uncertain variables, and wind power fuzzy parameters , photovoltaic fuzzy parameters and load fuzzy parameters need to be introduced.

[0108] The present application only considers the prediction error of wind power, photovoltaic power and load, and the strict power equality balance constraint is:

[0109] (31)

[0110] wherein, is the load prediction value of the system at t time; is the prediction error of the load at t time; is the wind power prediction value at t time; is the photovoltaic prediction value at t time; is the prediction error of the wind power at t time; is the prediction error of the photovoltaic at t time; is the wind curtailment amount of the wind power in the day-ahead scheduling plan at t time; is the photovoltaic curtailment amount of the photovoltaic in the day-ahead scheduling plan at t time; is the output of the conventional thermal power unit i at t time; is the maximum output of the conventional thermal power unit i; N G is the number of thermal power units.

[0111] Since the system power balance constraint contains uncertain variables, the power balance under the determined condition will no longer be applicable, so when arranging the day-ahead scheduling output of the system, the influence of the uncertainty factors of the system should be considered. In this paper, a fuzzy processing method is adopted, and wind power fuzzy parameters , photovoltaic fuzzy parameters and load fuzzy parameters Relaxing the constraints of formula (32), (33) to a certain confidence level The power balance under the condition of the balance constraint is established, and the possibility is not less than In this way, the set of uncertain factors is constructed:

[0112] (32)

[0113] (33)

[0114] wherein, Indicates the credibility of the event.

[0115] The optimization process of formula (32), (33) is to provide backup for the uncertainty of load, photovoltaic and wind power through conventional thermal power, so that the probability of balancing the total power supply in the system with the electrical load reaches the acceptable level of the dispatch decision maker. Compared with the deterministic constraint, the system credibility chance constraint has a processing mechanism for uncertain factors. Under this framework, the output of the conventional thermal power unit itself covers the backup capacity, so it is not necessary to set an independent backup power variable. The uncertainty in the system (such as prediction error) can be modeled by fuzzy parameters; the inequality constraint containing fuzzy parameters of the conventional unit output is regarded as an event, and the possibility of the event being established under uncertainty is not less than the preset confidence level. Such modeling method constitutes a fuzzy chance constrained programming problem.

[0116] The idea of fuzzy chance constraint for scheduling problem modeling is the model of fuzzy chance constrained programming for solving uncertainty problem: allow the scheduling result to not meet the constraint condition to some extent, but require the possibility of the scheduling result to be established to be not less than the confidence level set in advance by the scheduling decision maker.

[0117] The single-objective chance programming model with fuzzy variables is:

[0118] (34)

[0119] wherein, is the decision vector; is the parameter vector; is the objective function; is the constraint function; is the confidence level of the system.

[0120] In the process of solving fuzzy chance-constrained programming, the handling of chance-constrained is the key. There are two methods: for simple problems, the fuzzy parameters in the constraint conditions can be separated from the decision variables, or there is a certain linear relationship between the two, which can be converted into a clear equivalent class for processing, and then the traditional solving method is used for calculation; for more complex problems, the stochastic simulation technique is used. Simulation is an approximate process, the result is not accurate, and the size of the sample capacity is not easy to grasp. This paper uses the first method to handle fuzzy chance-constrained programming and its equivalent class processing method. The safety requirement of the system is very high, so the confidence level should not be too low, when ≥1 / 2, the clear equivalent class of chance-constrained is:

[0121] (35)

[0122] where, , are two assumed functions; is a part of function ; ~ (k=1,2,...,t, t∈R, R is the set of real numbers) is the membership degree parameter.

[0123] The expression of trapezoidal fuzzy parameter is:

[0124] (36)

[0125] where, is the fuzzy parameter in the system; is the predicted value; r1-r4 are the membership degree parameters of wind power and load in each period; w1-w4 are the proportionality parameters, which are generally determined by the historical data of fuzzy parameters.

[0126] Equation (32) is converted into the corresponding clear equivalent class according to the conversion method of trapezoidal parameters. This paper adopts the clear equivalent class under triangular fuzzy parameters.

[0127] The clear equivalent class of power balance is:

[0128] (37)

[0129] where, , are the membership degree parameters of wind power; , are the membership degree parameters of load.

[0130] Further, the step 3 specifically includes:

[0131] Step 2 introduces fuzzy parameters into the constraint conditions of the wind-solar-water-storage multi-energy complementary coordinated optimization scheduling model established in step 1 with the optimization objectives of minimum net load fluctuation, maximum operation benefit of the energy storage system, minimum operation cost of the thermal power unit and maximum renewable energy consumption capacity of wind, light and water, to form a wind-solar-water-storage multi-energy complementary coordinated optimization scheduling model with fuzzy parameters.

[0132] The upper scheduling model utilizes the rapid throughput capacity of the energy storage to follow the fluctuations of wind power, photovoltaic and load, and optimizes the output of the wind-solar-water-storage combined system with the objectives of minimum net load fluctuation and maximum operation benefit of the energy storage system, so as to reduce the peak load shifting pressure of the thermal power unit on the remaining load, that is, by optimizing the output of the energy storage system, the effect of the wind-solar-water-storage complementary system output tracking the load curve is optimal. The upper model uses the MATLAB platform to call the Cplex solver to solve the model, and obtains the optimal output of the wind-solar-water-storage combined system with minimum net load fluctuation and maximum operation benefit of the energy storage system. According to the output of the energy storage system in the combined system combined with the load prediction curve, the equivalent load curve is obtained, and the actual wind power consumption amount needs to be determined according to the peak shaving of the thermal power unit in the combined system.

[0133] The lower model passes the equivalent load curve transmitted by the upper model, takes the minimum operation cost of the thermal power unit and the minimum amount of abandoned wind, light and water as the optimization objectives, combines the abandoned wind and light penalty cost, peak shaving cost and the operation characteristics of each thermal unit, adopts the multi-objective pelican optimization algorithm for solving, determines the optimal economic scheduling strategy, then adopts the alternating iteration solving method, carries out the system peak shaving initiative verification, and ends the iteration after the peak shaving initiative constraint is satisfied, determines the thermal power unit participating in the deep peak shaving, and outputs the final scheduling strategy.

[0134] Multi-objective Pelican Optimization Algorithm (MPOA) is a newly developed population-based meta-heuristic optimization method, which is suitable for solving continuous multi-objective optimization problems. The algorithm has the advantages of fast convergence, few adjustable parameters, and simple computational structure. Its design inspiration comes from the group cooperation and individual hunting behavior exhibited by pelicans in the natural hunting process. Pelicans mainly feed on fish and have excellent swimming and flying abilities. Some individuals can circle at an altitude of 10-20 meters, lock the prey with sharp vision, and then dive to catch the prey. In more cases, pelicans use group coordination strategies to surround the fish school-they often form a straight line or U-shaped array, flap their wings to drive the fish school to the shallow water area, and then catch the fish collectively. Multi-objective Pelican Optimization Algorithm (MPOA) is based on the above biological behavior mechanism, and builds a corresponding mathematical model to simulate the search, cooperation and hunting process of pelicans, so as to effectively solve complex optimization problems.

[0135] (1) Exploration phase:

[0136] In this phase, the pelican determines the position of the prey and then rushes to the position of the prey. In the algorithm, the position of the prey is randomly generated in the search space. According to this mechanism, the exploration strategy of the pelican is modeled, so that the algorithm has the ability to explore different search areas. The update of the position of the pelican in each iteration is shown in equation (38).

[0137] (38)

[0138] where, is the new position of the i-th pelican in the j-th dimension in the current phase, is the position of the prey in the j-th dimension, I is a random number with a value of 1 or 2, is the objective function value of the lower model, represents the value of the fitness function of the i-th pelican.

[0139] In this phase, if the value of the fitness function is improved at the new position of the pelican, the new position is accepted and replaces the original position. This type of update can prevent the algorithm from moving to a non-optimal region. Equation (39) is used to model this process.

[0140] (39)

[0141] where, is the new position of the i-th pelican, is the objective function value of the i-th pelican based on the current phase.

[0142] (2) Development phase:

[0143] In this phase, the pelicans flap their wings on the water surface, thus forcing the fish to swim into the shallow water area, and then scoop them up into their mouths. The multi-objective pelican optimization algorithm is also based on this efficient hunting strategy, and this hunting behavior allows the hunting area of the algorithm to converge to a better point. This process improves the local search ability and development ability of the algorithm. This behavior of the pelicans during hunting is mathematically modeled in equation (40).

[0144] (40)

[0145] where, is the new position of the i-th pelican in the j-th dimension of this phase, R is a constant with a value of 0.2, is the neighborhood radius of , t is the iteration number, and T is the maximum iteration number. The coefficient represents the neighborhood radius of the population members, and this coefficient has a great influence on the development ability of the algorithm, which can help it to be closer to the global optimal solution. This is reflected in the fact that each member in the population performs local search within the neighborhood radius, thus converging to a better solution. In the early stage of iteration, the value of this coefficient is larger, accompanied by a larger search neighborhood radius of the population members. As the iteration number of the algorithm increases, the value of this coefficient decreases, and the search neighborhood of each member becomes smaller. This makes the algorithm more effectively search the area around each member with smaller and more accurate steps, so that it can converge to a solution closer to the global optimal solution.

[0146] In this phase, the updated position also needs to be verified by the fitness function. If the fitness of the new position is improved, the population member accepts the new position, otherwise it is rejected, which is modeled in equation (41).

[0147] (41)

[0148] where, is the new position of the i-th pelican, is the objective function value of the i-th pelican based on this phase. At the beginning of the search process, the algorithm creates a set of random initial solutions and calculates the fitness of each member. The pelicans perform position update iterations according to equations (38)-(41), and finally obtain the position of the currently best pelican.

[0149] Further, the Levy flight mechanism is added to the multi-objective pelican optimization algorithm. The Levy flight is a random walk strategy derived from the foraging behavior of various organisms in nature, such as insects, birds, and marine organisms. Its theoretical basis is the optimal search theory. This strategy is characterized by a power-law distribution of step size, which can achieve the organic combination of long and short distance movements in the search process. In the multi-objective pelican optimization algorithm, the Levy flight mechanism is introduced to improve the position updating method of the pelican individual: through its unique random step size distribution, the algorithm can effectively jump out of the local optimal region in a few long-distance jumps, thereby enhancing the global exploration ability; at the same time, when the step size is small, it helps fine search and improves the local development performance. Therefore, the introduction of Levy flight not only expands the search range of the algorithm, but also balances the ability between global exploration and local optimization.

[0150] The Levy flight strategy is introduced to update the pelican position, and equation (38) is adjusted as follows:

[0151] (42)

[0152] wherein, is a random step size based on Levy distribution, and the multiplication of and simulates the movement of prey in the form of Levy flight, is an adaptive parameter; The calculation formula is:

[0153] (43)

[0154] wherein Levy is the Levy flight function, and the calculation formula is:

[0155] (44)

[0156] wherein, is a fixed step size, taking the value of 1.5, and v are the standard deviations in two normal distributions and as follows:

[0157] (45)

[0158] For equation (42), the adaptive parameter is used to replace the random number rand in the original search strategy, wherein the calculation formula of E is as follows:

[0159] (46)

[0160] wherein c is a constant, taking the value of 0.5, t is the current iteration number, and T is the maximum iteration number.

[0161] The application of the technical scheme can be known, aiming at the strong uncertainty problem of renewable energy output such as wind power and photovoltaic and load demand, the application introduces fuzzy parameters to relax the power balance constraint, and constructs an opportunity constraint model based on the confidence level, which effectively expands the traditional deterministic scheduling framework to a robust decision system in an uncertain environment. This method not only avoids the scheduling mismatch risk caused by prediction error, but also retains the solvability of the model through the clear equivalent class transformation of fuzzy opportunity constraint, significantly improves the adaptability and safety of the scheduling scheme in actual operation; The application designs a hierarchical optimization scheduling framework, which minimizes the net load fluctuation and maximizes the operation benefit of the energy storage in the upper model, and minimizes the operation cost of the thermal power unit and the renewable energy curtailment in the lower model, realizes the organic coordination of "peak clipping and valley filling" and "economic scheduling". The upper model fully utilizes the rapid response characteristics of the energy storage system to smooth the wind and light output fluctuation, reduces the net load fluctuation amplitude borne by the thermal power unit, thereby reducing the equipment loss and fuel consumption caused by frequent start-stop and deep peak regulation; The lower model is based on the equivalent load curve, and the improved multi-objective pelican optimization algorithm (combined with Levy flight mechanism) is used to efficiently solve the coordinated output strategy of thermal power and renewable energy, maximize clean energy consumption under the premise of ensuring system safety. This hierarchical-collaborative mechanism takes into account the flexibility and economy of the system, effectively alleviating the peak regulation pressure brought by high proportion of renewable energy access; At the algorithm level, the improved multi-objective pelican optimization algorithm combines the biological group cooperative hunting behavior and the random search characteristics of Levy flight, which not only enhances the global exploration ability to avoid falling into local optimum, but also improves the local development precision through the adaptive neighborhood radius mechanism, which is suitable for solving high-dimensional, non-convex, multi-constrained complex scheduling problems. Compared with traditional intelligent algorithms, this method has advantages in convergence speed, solution set diversity and calculation stability, and provides a feasible tool for real-time or near-real-time scheduling of large-scale multi-energy systems. In summary, the application integrates uncertainty modeling, hierarchical collaborative optimization, multi-agent benefit coordination and advanced intelligent algorithms, which not only effectively improves the overall peak regulation capacity and renewable energy consumption level of the wind-solar-water-storage system, but also significantly reduces the system operation cost and the adjustment burden of the thermal power unit, enhances the operation flexibility, economy and reliability of the power system under the dual uncertainty of source and load, and provides a technical path with theoretical value and engineering practicality for building a high proportion of renewable energy power system.

[0162] According to a second aspect of the embodiments of the present application, there is provided a wind-solar-hydro-storage multi-energy system complementary peak regulation scheduling system considering source and load uncertainty, comprising the modules of the method in any of the above. The modules in the wind-solar-hydro-storage multi-energy system complementary peak regulation scheduling system considering source and load uncertainty can be realized by software, hardware and combinations thereof, in whole or in part. The modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the modules.

[0163] According to a third aspect of the embodiments of the present application, there is provided a terminal device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to perform the steps of the method in any of the above.

[0164] The specific embodiments of the present application are described in detail above with reference to the accompanying drawings, but the present application is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the purpose of the present application.

Claims

1. A wind-solar-hydro storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty, characterized in that, The method comprises the following steps: Step 1, establishing a wind-solar-water-storage multi-energy system complementary coordination optimization scheduling model with the minimum net load fluctuation, the maximum operation benefit and the minimum cost of the energy storage system, the minimum operation cost of the thermal power unit and the maximum renewable energy consumption capacity, and constructing constraint conditions for the wind-solar-water-storage multi-energy system complementary coordination optimization scheduling model; wherein the optimization objectives of the minimum net load fluctuation, the maximum operation benefit and the minimum cost of the energy storage system belong to the upper model, and the optimization objectives of the minimum operation cost of the thermal power unit and the maximum renewable energy consumption capacity belong to the lower model; the constraint conditions comprise power balance constraints, thermal power unit constraints, wind power output constraints, hydropower station output constraints, photovoltaic power station output constraints, energy storage constraints and line transmission capacity constraints; Step 2, rewriting the power balance constraint in the constraint conditions as a strict power equality balance constraint, introducing wind power fuzzy parameters, photovoltaic fuzzy parameters and load fuzzy parameters, relaxing the strict power balance constraint into a power constraint balance under a predetermined confidence level, so that the possibility of the balance constraint condition being established is not less than the predetermined confidence level, and thus constructing an uncertain factor set; Step 3, adopting a hierarchical optimization scheduling scheme for the wind-solar-water-storage multi-energy system complementary coordination optimization scheduling model with the introduced fuzzy parameters, solving the upper model by a solver, and solving the lower model by an optimization algorithm.

2. The wind-solar-hydro storage multi-energy system complementary peak regulation scheduling method considering source load uncertainty according to claim 1, characterized in that, The minimum net load fluctuation objective function expression is as follows: ； In the formula: is a dispatch cycle; is a net load value at time t; is an average value of the net load in a dispatch cycle. 3.The method of claim 1, wherein, The maximum operation benefit and minimum cost objective function expression of the energy storage system is as follows: ； In the formula: is the operating revenue of the energy storage system; is the operating cost of the energy storage system. 4.The method of claim 1, wherein, The minimum operation cost objective function expression of the thermal power unit is as follows: ； In the formula: is the cost of conventional peak regulation operation of the thermal power unit; is the total number of thermal power units; is the thermal power unit is the loss cost; is the output of the t-th thermal power unit at time t; is the oil injection cost of the t-th thermal power unit; is a dispatching period; and are the minimum technical output and the maximum output of the thermal power unit, respectively; is the stable combustion load value of the unit oil injection depth peak regulation; is the stable combustion load value of the unit non-oil injection depth peak regulation; is the output of the t-th thermal power unit at time t; is the output of the t-th thermal power unit at time t.

5. The method of claim 1, wherein, The maximum renewable energy consumption capacity expression is as follows: ； In the formula: is a scheduling period; is the total number of wind farms; is the total number of hydropower stations; is the total number of photovoltaic power stations; represents the wind farm the curtailed wind power occurring at time period t; is the photovoltaic power station the curtailed light power occurring at time period t; is the hydropower station the curtailed water power occurring at time period t; is the length of each time period. 6.The method of claim 1, wherein, The upper model is solved by a solver Cplex, and the lower model is solved by a multi-objective pelican optimization algorithm.

7. A wind-solar-hydro storage multi-energy system complementary peak regulation scheduling system considering source-load uncertainty, characterized in that, The module comprises the method in any one of claims 1-6.

8. A terminal device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, The processor is configured to perform the steps of the method in any one of claims 1-6.

Images