A calculation method for the optimal dispatching model of a hydro-wind-solar hybrid power generation system

Abstract

The present invention provides a calculation method for the optimization scheduling model of the water-wind-solar complementary power generation system. By analyzing the output characteristics of water, wind and solar, the characterization method and evaluation index of the water-wind-solar complementarity are further studied. The present invention establishes a model with the minimum standard deviation of the residual value of the combined system output as the objective function. At the same time, constraints that can maintain the stable operation of the system and parameters that conform to the actual operating conditions of each unit are set for each part of the model. Then, based on the traditional particle swarm algorithm, a particle swarm algorithm in which the inertia weight changes adaptively with the number of iterations is proposed. The characteristics of the improved algorithm and its advantages over the algorithm before the improvement are further analyzed. Its calculation process is further explained, and it is prepared to use the improved particle swarm algorithm to optimize and solve the combined system.

Description

Technical Field

[0001] The present invention relates to the technical field of hydraulic turbine control, and in particular to a calculation method for an optimization scheduling model of a water-wind-solar complementary power generation system. Background Art

[0002] As society develops, new demands are being placed on energy, and the development of high-quality clean energy is becoming a trend in current social energy development. The coordinated operation of multiple energy sources is crucial to improving the integration of clean energy sources such as wind power and photovoltaics into the power system, while also ensuring the smooth operation of the entire power system. Research into the optimal scheduling of multiple clean energy sources, combined with complementary resources, is imperative to address environmental and energy pressures.

[0003] At present, the dispatch model of the combined power generation system based on economic dispatch is mainly used. This dispatch model ignores the further exploration of the correlation between wind and solar output, and does not further consider the connection between wind turbines and photovoltaic power plants. In the modeling process, they are simply treated as independent power sources. Summary of the Invention

[0004] To address the above issues, the present invention provides a calculation method for an optimized scheduling model for a hydropower-wind-solar hybrid power generation system. By analyzing the output characteristics of hydropower, wind, and solar power, further research is conducted on the characterization and evaluation indicators of hydropower-wind-solar hybrid power generation. A model is established with the minimum standard deviation of the combined system output residual value as the objective function. Constraints that maintain stable system operation and parameters that conform to the actual operating conditions of each unit are set for each component of the model. Then, based on the traditional particle swarm algorithm, a particle swarm algorithm is proposed in which the inertia weight adaptively changes with the number of iterations. The characteristics of the improved algorithm and its advantages over the original algorithm are further analyzed. The computational process is further elaborated, and the improved particle swarm algorithm is used to optimize and solve the combined system.

[0005] In order to achieve the above technical features, the purpose of the present invention is to achieve the following: a calculation method for an optimal scheduling model of a hydro-wind-solar complementary power generation system:

[0006] Step 1: Establish an optimized scheduling model:

[0007] The objective function is to minimize the residual load, which is also a related indicator of output fluctuation, and minimize the standard deviation of the system output residual value, and to establish the optimal scheduling model accordingly.

[0008] Step 2: Construct the objective function:

[0009] Based on the fluctuating nature of wind and solar power output, the optimization scheduling model uses the standard deviation of the output margin in each period and the average output margin as the objective function;

[0010] Step 3: Determine the constraints:

[0011] Step 4: Solve the objective function based on the improved particle swarm algorithm:

[0012] Based on the traditional particle swarm optimization algorithm, an improved particle swarm optimization algorithm in which the inertia weight changes adaptively with the number of iterations is proposed. The improved particle swarm optimization algorithm is used to optimize and solve the objective function.

[0013] Preferably, during the establishment of the optimization scheduling model in step 1, the process of exerting the peak-shaving capacity of hydropower will be affected by the volatility and anti-peak-shaving characteristics of the load level and the wind and solar grid connection. Based on this, a comprehensive optimization scheduling model is constructed to solve the amplitude modulation problem of the hydropower station when the wind and solar power are coordinated.

[0014] Preferably, in the short-term coordinated optimization scheduling of wind, solar and water in step 1, the peak-shaving problem needs to be given priority consideration in the process of wind and solar joint grid connection. In the wind and solar complementary energy combination based on hydropower stations, the rapid regulation capability of hydropower is utilized to cooperate with the joint operation of wind power and photovoltaics.

[0015] Preferably, the objective function established in step 2 is:

[0016] ;

[0017] ;

[0018] Where, for t The residual value of the combined system output during the time period; for t Total load of the combined system during the time period; 、 、 Wind, solar and water resources are t Output during the time period; For the total period, is the average value of the residual value of the combined system output; is the objective function.

[0019] Preferably, the constraints in step 3 include:

[0020] (1) Wind and solar power output constraints:

[0021] Taking into account the operating costs of various power sources, it is desirable to maximize the output of wind and photovoltaic power when operating a combined system. However, wind and solar generators have their own output limitations that need to be taken into account. This is primarily due to their rated power, which determines their upper output limit.

[0022] ;

[0023] ;

[0024] Where, 、 are the maximum outputs of wind power and photovoltaic power respectively;

[0025] (2) Wind and solar power consumption constraints:

[0026] Excessive wind and solar power absorption will lead to unstable system operation, so the corresponding absorption power must be ensured to be within the upper and lower limits that the combined system can withstand;

[0027] ;

[0028] Where, 、 are the minimum and maximum values of the combined system's power absorption capacity, respectively;

[0029] (3) Hydropower conversion relationship:

[0030] ;

[0031] Where, is the power generation efficiency of the turbine unit, which is a constant; Current t flow and head at any given moment;

[0032] (4) Grid-connected balance:

[0033] The grid-connected power is the sum of hydropower and wind and solar power output;

[0034] ;

[0035] Where, is the grid-connected power of the combined system;

[0036] (5) Hydropower output constraints:

[0037] ;

[0038] Where, 、 are the lower and upper limits of hydropower output, Contribute to hydropower;

[0039] (6) Outbound flow constraints:

[0040] ;

[0041] Where, for t The outflow of hydropower at the moment, 、 They are the lower limit and upper limit of outbound flow respectively;

[0042] (7) Reservoir water level constraints:

[0043] ;

[0044] Where, for t The reservoir water level at the moment, 、 They are the lower limit and upper limit of the reservoir water level respectively.

[0045] Preferably, the specific process of the particle swarm algorithm is as follows:

[0046] Assume that the population is in a D-dimensional space, where there are M particles, defined as , where S is the space set, , , in this space The solution contains information about the position and velocity of each particle;

[0047] The position formula is: , ,in Indicates the upper and lower limits of the particle's position change during motion;

[0048] The speed formula is: ,in , Respectively represent the maximum and minimum values of particle velocity during the iteration process;

[0049] Global optimal position: ;

[0050] Individual optimal position: ;

[0051] The particle velocity update formula is:

[0052] ;

[0053] The particle position update formula is:

[0054] ;

[0055] Where: is the inertia weight, which inherits the past velocity of the particle and represents the inertial characteristics of the particle; , is the learning factor, which ranges from 1 to 2. It represents the individual cognition in the particle swarm algorithm, which means that during the iteration process, the particle needs to compare its previous value to achieve the position change of the particle; It reflects public knowledge, which means that during the iteration process, each particle needs to be compared with the best position currently recognized; 、 It is a random number, which is assigned a random number between 0 and 1 as the initial value through a random function. 、 The assignment of is completely random, independent of each other and do not affect each other.

[0056] Preferably, the fitness value is used to evaluate the quality of particles and is set as the objective function value. In addition, a large inertia weight is beneficial to global search, whereas a small inertia weight is beneficial to local search.

[0057] If the inertia weight is kept fixed, the details of the problem solution will change as the iterative process progresses, resulting in computational defects. Based on this, the inertia weight that changes with the iterative process is introduced to achieve dynamic adaptation of the entire solution process, thereby obtaining an improved particle swarm algorithm.

[0058] There are two different adaptive inertia weight solutions:

[0059] When it is necessary to solve the minimum or maximum problem, the formula for recalculating the inertia weight w value in each iteration is as follows:

[0060] Minimum formula:

[0061] = ;

[0062] Maximum value formula:

[0063] = ;

[0064] Where: and are the preset minimum and maximum inertia coefficients; and are the average fitness of all particles at the dth iteration; It is the minimum or maximum fitness of all particles at the dth iteration; the smaller the fitness, the closer it is to the optimal solution, and local search is more needed at this time; the larger the fitness, the farther it is from the optimal solution, and global search is needed at this time.

[0065] Preferably, the Take 0.4, Take 0.9.

[0066] Preferably, the solution process in step 4 is based on the improved particle swarm algorithm:

[0067] Step 4.1, parameter setting:

[0068] Step 4.2, particle initialization:

[0069] Use random functions to initialize the initial position and velocity of the particles;

[0070] Step 4.3, solve the target value of all particles in each time period:

[0071] Determine the individual optimal value pbest of each particle and the optimal value Gbest of the entire group;

[0072] Step 4.4, based on the initial inertia weight data, use the speed and position formula of the particle swarm to calculate the updated speed, position and fitness of the particle in each iteration;

[0073] Step 4.5: Take out the fitness of each particle and calculate the average fitness of this iteration accordingly. and minimum fitness ;

[0074] Step 4.6: Update the inertia weight w using the fitness-related values calculated in step 4.5 and the minimum formula of the adaptive inertia weight, and continue the iterative process;

[0075] In step 4.7, the next step is performed based on whether the global optimum and the number of iterations meet the termination conditions; if so, Pbest and Gbest are output, otherwise steps 4.2 to 4.6 will continue to loop until the requirements are met.

[0076] Preferably, in step 4.1, the specific process of parameter setting is that it involves wind power, photovoltaic power stations, and hydropower stations, so the dimension is set to 3, the number of particles is 200, the particle movement range is the upper and lower limits of the actual operating output of each generator set, the number of iterations is 500, and the learning factor is 1. 、 Set to 2.

[0077] The present invention has the following beneficial effects:

[0078] This invention explores an optimal scheduling model for a combined wind, solar, and hydropower system. First, by describing the basic structure and structural principles of wind turbines, solar power generation equipment, and hydropower station generators, the feasibility of a wind-solar hybrid combined power generation system for river basin hydropower stations is further analyzed. Furthermore, the optimal scheduling of this system is studied. The algorithm provided by this invention uses the minimum standard deviation of the combined system output and load as the objective function. Taking operational stability and power balance into account, the algorithm combines the basic constraints of wind power, photovoltaic power, and hydropower, the predicted compliance of a typical day in a specific region, and the output constraints of each energy source. A combined wind-solar hybrid power generation system for a hydropower station is established, and optimal scheduling simulations and further analysis are performed. To solve the optimal scheduling problem of the objective function, this invention proposes an improved particle swarm algorithm based on adaptive inertia weighting to solve the optimal scheduling problem of the combined wind, solar, and hydropower system. Based on the principles and operational flow of the particle swarm algorithm, the shortcomings of the basic algorithm in terms of convergence speed, applicability to specific models, and optimal selection are analyzed, and a new improved method is proposed. This invention proposes an adaptive inertia weighted particle swarm algorithm. A specific example simulation of optimal scheduling for a typical day in a specific region is implemented using MATLAB programming, ultimately yielding the optimal solution. The research results prove that the improved algorithm for optimizing the output of each part of the water-wind-solar complementary system in the basin can better ensure the stable operation of the joint system. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] The present invention will be further described below with reference to the accompanying drawings and examples.

[0080] Figure 1 The calculation process of PSO is improved for the present invention.

[0081] Figure 2 This is the typical daily load forecast curve in Example 2 of this invention.

[0082] Figure 3 This is the maximum output curve of the wind power plant in each period in Example 2 of the present invention.

[0083] Figure 4 This is the change in storage capacity of the hydropower station in each period in Example 2 of this invention.

[0084] Figure 5 This is the output curve of each part of the combined system in Example 2 of the present invention. DETAILED DESCRIPTION

[0085] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0086] Example 1:

[0087] By analyzing the output characteristics of hydropower, wind power, and solar power, further research was conducted on the characterization and evaluation indicators of hydropower, wind power, and solar power complementarity. This paper established a model with the objective function of minimizing the standard deviation of the residual output of the combined system. Constraints were set for each component of the model to maintain stable system operation, as well as parameters that conform to the actual operating conditions of each unit. Then, based on the traditional particle swarm algorithm, a particle swarm algorithm was proposed in which the inertia weight adaptively changes with the number of iterations. The characteristics of the improved algorithm and its advantages over the original algorithm were further analyzed. The computational process was further elaborated, and the improved particle swarm algorithm was used to optimize and solve the combined system.

[0088] This invention uses the total power output of electricity as its fundamental data, comprehensively considering various factors of power quality, including output stability, to establish a mathematical model for coordinated scheduling of multiple energy sources. In the short-term coordinated optimization of wind, solar, and hydropower, peak shaving is a key consideration during the wind-solar grid connection process. In a wind-solar complementary energy combination based on a hydropower station, the rapid regulation capabilities of hydropower are utilized to support the combined operation of wind and photovoltaic power.

[0089] The process of exerting the peak-shaving capacity of hydropower is often affected by the volatility and anti-peak-shaving characteristics of the load level and wind and solar grid connection. Therefore, a comprehensive model can be constructed to effectively solve the amplitude modulation problem of the hydropower station when the wind and solar power are coordinated. Based on the above theoretical basis, the present invention selects the minimum residual load, that is, the relevant index of output volatility, and the minimum standard deviation of the residual value of the combined system output as the objective function, and thereby realizes the establishment of the optimization scheduling model. The objective function of the residual load is constructed based on the data of the comprehensive output of the combined system and the predicted load of the day. Make good use of the peak-shaving capacity of the hydropower station, exert the corresponding peak-shaving effect, and observe whether the combined output is stable through the volatility of the residual output. Based on the above theoretical analysis, the objective function and corresponding constraints established by the present invention are specifically introduced as follows.

[0090] Objective function:

[0091] ;

[0092] ;

[0093] Where, for t The residual value of the combined system output during the time period; for t Total load of the combined system during the time period; 、 、 Wind, solar and water resources are t Output during the time period; For the total period, is the average value of the residual value of the combined system output; as the objective function.

[0094] Constraints:

[0095] (1) Wind and solar power output constraints:

[0096] Taking into account the operating costs of various power sources, it is desirable to maximize the output of wind and photovoltaic power when operating a combined system. However, wind and solar generators have their own output limitations that need to be taken into account. This is primarily due to their rated power, which determines their upper output limit.

[0097] ;

[0098] ;

[0099] Where, 、 are the maximum outputs of wind power and photovoltaic power respectively;

[0100] (2) Wind and solar power consumption constraints:

[0101] Excessive wind and solar power absorption will lead to unstable system operation, so the corresponding absorption power must be ensured to be within the upper and lower limits that the combined system can withstand;

[0102] ;

[0103] Where, 、 are the minimum and maximum values of the combined system's power absorption capacity, respectively;

[0104] (3) Hydropower conversion relationship:

[0105] ;

[0106] Where, is the power generation efficiency of the turbine unit, which is a constant, generally 0.81; Current t flow and head at any given moment;

[0107] (4) Grid-connected balance:

[0108] The grid-connected power is the sum of hydropower and wind and solar power output;

[0109] ;

[0110] Where, is the grid-connected power of the combined system;

[0111] (5) Hydropower output constraints:

[0112] ;

[0113] Where, 、 Lower and upper limits of hydropower output; Contribute to hydropower;

[0114] (6) Outbound flow constraints:

[0115] ;

[0116] Where, for t The outflow of hydropower at the moment, 、 They are the lower limit and upper limit of outbound flow respectively.

[0117] (7) Reservoir water level constraints:

[0118] ;

[0119] Where, for t The reservoir water level at the moment, 、 They are the lower limit and upper limit of the reservoir water level respectively.

[0120] The basic principle of particle swarm optimization (PSO) algorithm:

[0121] Consider a space containing several particles to be solved. These particles represent the solutions to the problem being solved. A random function is then used to randomly assign acceptable initial values to these particles. In other words, each particle is assigned an acceptable initial position and velocity. The next step is to incorporate these particle positions into the model formula for the first solution. The solution results are used to assess the quality of each position, resulting in both individual and global optima. The optimal values are then iterated using the updated position and velocity calculation formulas. The resulting values are then applied to the objective function for the next iteration and then solved. The results are compared with the previous values, and both values are updated. After each iteration, the optimality is reevaluated to determine whether they meet the termination criteria. If so, the iteration stops; otherwise, the iteration continues until the termination criteria are met. Following this process, the resulting Pbest and Gbest values are considered to be the solutions that meet the criteria.

[0122] For example, suppose the population is in a D-dimensional space with M particles, defined as , where S is the space set, , , in this space The solution contains information about the position and velocity of each particle;

[0123] The position formula is: , ,in Indicates the upper and lower limits of the particle's position change during motion;

[0124] The speed formula is: ,in , Respectively represent the maximum and minimum values of particle velocity during the iteration process;

[0125] Global optimal position: ;

[0126] Individual optimal position: ;

[0127] The particle velocity update formula is:

[0128] ;

[0129] The particle position update formula is:

[0130] ;

[0131] Where: is the inertia weight, which inherits the past velocity of the particle and represents the inertial characteristics of the particle; , It is the learning factor, which ranges from 1 to 2. In commonly used optimization algorithms, the value is generally 2. It represents the individual cognition in the particle swarm algorithm, which means that during the iteration process, the particle needs to compare its previous value to achieve the position change of the particle; It reflects public knowledge, which means that during the iteration process, each particle needs to be compared with the best position currently recognized; 、 It is a random number, which is assigned a random number between 0 and 1 as the initial value through a random function. 、 The assignment of is completely random, independent of each other and do not affect each other.

[0132] Improved PSO-adaptive inertia weight:

[0133] According to the basic introduction of PSO, we know that the fitness value is used to evaluate the quality of particles and is generally set as the objective function value. In addition, we know that a larger inertia weight is conducive to global search, while a smaller inertia weight is conducive to local search.

[0134] The particle velocity iteration formula of basic PSO is:

[0135] ;

[0136] is the particle's own inertia weight. If this inertia weight remains fixed, the details of the problem solution will change as the iterative process progresses, exposing many computational flaws. To address these issues, the present invention considers introducing an inertia weight that changes continuously with the iterative process, enabling dynamic adaptation of the entire solution process.

[0137] Below are two examples of solving problems using adaptive inertia weights.

[0138] When it is necessary to solve the minimum or maximum problem, the formula for recalculating the inertia weight w value in each iteration is as follows:

[0139] Minimum formula:

[0140] = ;

[0141] Maximum value formula:

[0142] = ;

[0143] Where: and are the preset minimum and maximum inertia coefficients; and are the average fitness of all particles at the dth iteration; It is the minimum or maximum fitness of all particles at the dth iteration; the smaller the fitness, the closer it is to the optimal solution, and local search is more needed at this time; the larger the fitness, the farther it is from the optimal solution, and global search is needed at this time.

[0144] Compared to the basic particle swarm algorithm, the inertia weight is now related to each iteration and the fitness of each particle. This formula can be understood as follows: because the problem being solved is the maximum value (maximum fitness), when a particle's fitness is less than the average fitness, it indicates that the particle is relatively far from the maximum value, and the search range needs to be expanded to find the maximum value. When a particle's fitness is greater than the average fitness, it indicates that the particle is close to the maximum value, and the search range needs to be narrowed for a local, precise search.

[0145] Calculation process based on improved PSO:

[0146] See also Figure 1 ,Based on ,PSO, the solution process based on adaptive inertia weight PSO is as follows:

[0147] Step 4.1, parameter setting. Involving wind power, photovoltaic, and hydropower stations, the dimension is set to 3, the number of particles is 200, the particle movement range is the upper and lower limits of the actual operating output of each generator set, the number of iterations is 500, and the learning factor is 1. 、 Set to 2.

[0148] Step 4.2, particle initialization. Use a random function to initialize the initial position and velocity of the particle;

[0149] Step 4.3: Solve for the target values of all particles in each time period. Determine the individual optimal value pbest for each particle and the optimal value Gbest for the entire group.

[0150] Step 4.4, based on the initial inertia weight data, use the speed and position formula of the particle swarm to calculate the updated speed, position and fitness of the particle in each iteration;

[0151] Step 4.5: Take out the fitness of each particle and calculate the average fitness of this iteration accordingly. and minimum fitness ;

[0152] Step 4.6: Update the inertia weight w using the fitness-related values calculated in step 4.5 and the minimum formula of the adaptive inertia weight, and continue the iterative process;

[0153] In step 4.7, the next step is performed based on whether the global optimum and the number of iterations meet the termination conditions; if so, Pbest and Gbest are output, otherwise steps 4.2 to 4.6 will continue to loop until the requirements are met.

[0154] Example 2:

[0155] To verify the applicability of the adaptive inertia weighted PSO algorithm proposed in this paper for optimizing the scheduling of a wind-solar hybrid system for river basin hydropower stations, the following example was used. The simulation model used a combined power generation system consisting of two river basin hydropower stations, one wind farm, and one photovoltaic power station, with a daily scheduling period of hourly intervals. The data used were basic data from a typical day in the region. The optimized scheduling solution was solved using an improved PSO algorithm based on a Matlab program to verify the effectiveness and applicability of the improved optimization method.

[0156] (1) Model parameter setting:

[0157] Table 1 Inflow rate of cascade hydropower station reservoirs (unit: m 3 / s)

[0158]

[0159] Table 2 Storage capacity and output upper and lower limits of cascade hydropower stations (unit: m 3 / s)

[0160]

[0161] Table 3 Wind turbine photovoltaic power supply operating parameters (unit: WM, yuan / KW)

[0162]

[0163] (II) Typical load curve and wind and solar output constraints:

[0164] The typical day data of hydropower stations in a certain area of the Jinsha River Basin were selected for scheduling optimization calculation.

[0165] Figure 2 shows the load curve for a typical day in the local area. This curve exhibits typical characteristics of the electricity load at hydropower stations in the basin. In the morning, load fluctuations increase as flow increases, with the maximum predicted load occurring around 10:00 p.m., reaching 115 MW. As electricity consumption decreases in the afternoon and into the evening, the load gradually decreases, dropping to around 50 MW by midnight.

[0166] like Figure 3 This is the maximum output curve for a wind power plant. Observing the wind turbine output curve, we can see that the wind turbine output fluctuates irregularly over time. On a typical day, the minimum output is 10 MW in the first hour, while the maximum output reaches 29.4 MW around 17:00. The output fluctuates between these two values at other times. Between 10:00 and 1:00 PM, as wind speed drops sharply, the wind turbine output also decreases accordingly. At noon, the wind turbine output is approximately 5 MW, and at 1:00 PM, it is only 11.3 MW.

[0167] Observing the output curve of a photovoltaic power plant reveals significant day-to-night differences. Around 1:00 PM, when sunlight is strongest, the photovoltaic power output reaches a maximum of 26 MW. However, from 8:00 PM onwards and before 6:00 AM, during the night, when sunlight is almost non-existent, the photovoltaic power generation system produces virtually no output.

[0168] Comparing the wind turbine and photovoltaic output curves in the figure, we can see that even at night, despite the lack of sunlight and no photovoltaic output, wind turbines can still maintain a certain power output. This is an important basis for their complementarity in time and space.

[0169] (III) Analysis of typical day results:

[0170] The present invention selects the basic data of a typical day in the region. The optimization scheduling cycle is 24 hours in a day. And the corresponding optimization configuration is carried out accordingly. Figure 3 A simulation was conducted based on the given output limits and the hydropower station data given in Tables 1 through 3. Programming was performed using the MATLAB platform to implement the aforementioned optimal scheduling process. While ensuring the operational stability of the combined system, the system's remaining output was maintained stable, and the total daily output power was maximized. Using the optimal scheduling model proposed in this invention, calculations yielded the data shown in Table 4.

[0171] Table 4 Output parameters of each part of the combined system at each time period (unit: h, WM)

[0172]

[0173] The data in Table 4 shows that the combined system reaches a maximum output of approximately 120 MW around 10:00 AM. Due to natural factors, wind power output fluctuates, reaching a maximum of 26.3 MW at 10:00 AM. Photovoltaic power output is affected by the time of day, with sunlight intensity near zero at night. It reaches its peak around noon, at approximately 23.0 MW. The total hydropower station output reaches its maximum at 9:00 AM, with a maximum output of approximately 73.6 MW.

[0174] like Figure 4 The graph shows the change in the total average reservoir capacity of the hydropower station over time during various time periods. It can be seen that from 0:00 to 5:00, the water inflow is high, and the corresponding reservoir capacity increases continuously. From 5:00 to 10:00, the hydropower station is in power generation mode, and the reservoir capacity decreases continuously. The total reservoir capacity of the hydropower station fluctuates steadily around 91.5 million cubic meters on a typical day.

[0175] Figure 4 This is the output curve of each component of the combined system within a 24-hour day, obtained by the optimization algorithm. The operating principle of the combined system is to ensure that the standard deviation of the system's output surplus is minimized, the total daily power generation is maximized, and the scheduling results meet the operating restrictions of each unit. Due to the unstable and unpredictable output of wind power and photovoltaic power, when the combined system output is sufficient, the present invention limits the grid-connected power of wind power and photovoltaic power. Generally speaking, due to factors such as the scale and output constraints of wind and solar power units, the output of wind power and photovoltaic power in the system is generally smaller than that of hydropower units.

Claims

1. A calculation method for an optimal scheduling model of a hydro-wind-solar hybrid power generation system, characterized by: Step 1: Establish an optimized scheduling model: The objective function is to minimize the residual load, which is also a related indicator of output fluctuation, and minimize the standard deviation of the system output residual value, and to establish the optimal scheduling model accordingly. Step 2: Construct the objective function: Based on the fluctuating nature of wind and solar power output, the optimization scheduling model uses the standard deviation of the output margin in each period and the average output margin as the objective function; Step 3: Determine the constraints: Step 4: Solve the objective function based on the improved particle swarm algorithm: Based on the traditional particle swarm algorithm, an improved particle swarm algorithm with adaptive changes of inertia weight with the number of iterations is proposed. The improved particle swarm algorithm is used to optimize the objective function. The objective function established in step 2 is: ； ； Where, for t The residual value of the combined system output during the time period; for t Total load of the combined system during the time period; 、 、 Wind, solar and water resources are t Output during the time period; For the total period, is the average value of the residual value of the combined system output; is the objective function; The constraints in step 3 include: (1) Wind and solar power output constraints: Taking into account the operating costs of various power sources, it is desirable to maximize the output of wind and photovoltaic power when operating a combined system. However, wind and solar generators have their own output limitations that need to be taken into account. This is primarily due to their rated power, which determines their upper output limit. ； ； Where, 、 are the maximum outputs of wind power and photovoltaic power respectively; (2) Wind and solar power consumption constraints: Excessive wind and solar power absorption will lead to unstable system operation, so the corresponding absorption power must be ensured to be within the upper and lower limits that the combined system can withstand; ； Where, 、 are the minimum and maximum values of the combined system's power absorption capacity, respectively; (3) Hydropower conversion relationship: ； Where, is the power generation efficiency of the turbine unit, which is a constant; Current t flow and head at any given moment; (4) Grid-connected balance: The grid-connected power is the sum of hydropower and wind and solar power output; ； Where, is the grid-connected power of the combined system; (5) Hydropower output constraints: ； Where, 、 are the lower and upper limits of hydropower output, Contribute to hydropower; (6) Outbound flow constraints: ； Where, for t The outflow of hydropower at the moment, 、 They are the lower limit and upper limit of outbound flow respectively; (7) Reservoir water level constraints: ； Where, for t The reservoir water level at the moment, 、 They are the lower limit and upper limit of the discharge water level respectively.

2. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 1 is characterized in that: During the establishment of the optimization scheduling model in step 1, the process of exerting the peak-shaving capacity of hydropower will be affected by the volatility and anti-peaking characteristics of the load level and the wind and solar grid connection. Based on this, a comprehensive optimization scheduling model is constructed to solve the amplitude modulation problem of the hydropower station when the wind and solar power are coordinated.

3. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 2 is characterized in that: In the short-term coordinated optimization scheduling of wind, solar and hydropower in step 1, the peak-shaving problem needs to be given priority consideration in the process of wind and solar power grid connection. In the wind and solar power complementary energy combination based on hydropower stations, the rapid regulation capability of hydropower is utilized to coordinate the joint operation of wind power and photovoltaic power.

4. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 3 is characterized in that: The specific process of the particle swarm algorithm is as follows: Assume that the population is in a D-dimensional space, where there are M particles, defined as , where S is the space set, , , in this space The solution contains information about the position and velocity of each particle; The position formula is: , ,in Indicates the upper and lower limits of the particle's position change during motion; The speed formula is: ,in , Respectively represent the maximum and minimum values of particle velocity during the iteration process; Global optimal position: ; Individual optimal position: ; The particle velocity update formula is: ； The particle position update formula is: ； Where: is the inertia weight, which inherits the past velocity of the particle and represents the inertial characteristics of the particle; , is the learning factor, which ranges from 1 to 2. It represents the individual cognition in the particle swarm algorithm, which means that during the iteration process, the particle needs to compare its previous value to achieve the position change of the particle; It reflects public knowledge, which means that during the iteration process, each particle needs to be compared with the best position currently recognized; 、 It is a random number, which is assigned a random number between 0 and 1 as the initial value through a random function. 、 The assignment of is completely random, independent of each other and do not affect each other.

5. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 4 is characterized in that: The fitness value is used to evaluate the quality of particles and is set as the objective function value. In addition, a large inertia weight is conducive to global search, while a small inertia weight is conducive to local search. If the inertia weight is kept fixed, the details of the problem solution will change as the iterative process progresses, resulting in computational defects. Based on this, the inertia weight that changes with the iterative process is introduced to achieve dynamic adaptation of the entire solution process, thereby obtaining an improved particle swarm algorithm. There are two different adaptive inertia weight solutions: When it is necessary to solve the minimum or maximum problem, the formula for recalculating the inertia weight w value in each iteration is as follows: Minimum formula: = ； Maximum value formula: = ； Where: and are the preset minimum and maximum inertia coefficients; and are the average fitness of all particles at the dth iteration; It is the minimum or maximum fitness of all particles at the dth iteration; the smaller the fitness, the closer it is to the optimal solution, and a local search is more needed at this time; the larger the fitness, the farther it is from the optimal solution, and a global search is needed at this time.

6. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 5 is characterized in that: described Take 0.4, Take 0.

9.

7. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 6 is characterized in that: The solution process in step 4 is based on the improved particle swarm algorithm: Step 4.1, parameter setting: Step 4.2, particle initialization: Use random functions to initialize the initial position and velocity of the particles; Step 4.3, solve the target value of all particles in each time period: Determine the individual optimal value pbest of each particle and the optimal value Gbest of the entire group; Step 4.4, based on the initial inertia weight data, use the speed and position formula of the particle swarm to calculate the updated speed, position and fitness of the particle in each iteration; Step 4.5: Take out the fitness of each particle and calculate the average fitness of this iteration accordingly. and minimum fitness ; Step 4.6: Update the inertia weight w using the fitness-related values calculated in step 4.5 and the minimum formula of the adaptive inertia weight, and continue the iterative process; In step 4.7, the next step is performed based on whether the global optimum and the number of iterations meet the termination conditions; if so, Pbest and Gbest are output, otherwise steps 4.2 to 4.6 will continue to loop until the requirements are met.

8. The calculation method of the optimal scheduling model of the hydro-wind-solar complementary power generation system according to claim 7 is characterized in that: In step 4.1, the specific process of parameter setting is as follows: wind power, photovoltaic power station, and hydropower station are involved, so the dimension is set to 3, the number of particles is 200, the particle movement range is the upper and lower limits of the actual operating output of each generator set, the number of iterations is 500, and the learning factor is 1. 、 Set to 2.

Images