Multi-source microgrid optimization scheduling method and system

Abstract

The invention relates to the technical field of energy optimization scheduling, in particular to a multi-source microgrid optimization scheduling method and system. The method comprises the following steps: acquiring a mathematical model of each distributed power supply and an energy conversion device of a micro-grid in a to-be-optimized region, and constructing a target function by taking weighted minimization of fuel cost and operation maintenance cost of the micro-grid, pollutant discharge treatment cost of the micro-grid and load control compensation cost of the micro-grid as a target; setting a power balance constraint, an energy storage charge state constraint, a generation power constraint, each distributed power supply output constraint and a ramp rate constraint as constraint conditions, and constructing a day-ahead optimization model; the day-ahead optimization model is solved through the whale-bat algorithm, and the whale-bat algorithm replaces the search stage of the whale algorithm with the improved bat algorithm to solve the day-ahead optimization scheduling strategy. The optimal scheduling strategy obtained by the invention can respond to dynamic changes more quickly while ensuring high precision.

Description

Technical Field

[0001] The present invention relates to the technical field of energy optimization scheduling, and in particular to a multi-source microgrid optimization scheduling method and system. Background Art

[0002] For a long time, large-scale centralized power generation and long-distance power transmission have been the primary methods to meet the growing demand for electricity, which has brought great convenience to people's lives, but also have many defects, such as huge investment, long construction period, poor operation flexibility, high risk of power outages, etc. Therefore, microgrids, which can supplement large power grids and help to effectively utilize renewable energy, have been increasingly selected and applied in countries around the world.

[0003] Existing multi-source microgrid optimization scheduling methods usually establish mathematical models of various energy resources in the microgrid (such as wind energy, solar energy, energy storage systems and generators), and set objective functions based on these models. The objective function takes into account factors such as the economic efficiency of different resources, the proportion of renewable energy and the stability of the system. In order to solve the optimization problem, the intelligent optimization algorithm usually solves the multi-source microgrid optimization scheduling strategy. Multi-source microgrids involve multiple energy sources, and their loads and supplies are highly uncertain. Compared with other intelligent algorithms, the whale algorithm simulates and imitates the predation methods of whales in the ocean when dealing with complex optimization problems. It simulates three predation mechanisms in the whale predation process. Its unique predation mechanism enables the whale algorithm to conduct effective random exploration and local refinement during the search process. Therefore, the global search capability of the whale algorithm is very good, and it can effectively explore the optimal microgrid optimization scheduling solution in a complex multi-source environment. In addition, in the dynamic environment of multi-source microgrids, resource supply and demand may change frequently. The whale algorithm has fewer parameters and can maintain robustness and stability in different load and environmental changes. It is relatively simple to ensure the effectiveness of the scheduling scheme and does not require complex parameter adjustments, making it easier to apply and adjust in different optimization tasks. Therefore, in the context of multi-source microgrid optimization scheduling, choosing the whale algorithm helps to achieve the optimization goal more efficiently.

[0004] However, although the global search capability of the whale algorithm is excellent and can effectively find the global optimal solution, its search phase updates individual positions by randomly selecting target positions. In a complex optimization space, randomly selecting target positions may cause the algorithm to repeatedly search near a local optimal solution and find it difficult to jump out of this range. Therefore, this method is prone to fall into local optimality in complex, multi-peak optimization problems. Especially in scheduling problems, as the iteration proceeds, the algorithm is prone to lose diversity, which makes it easier for the algorithm to fall into similar solutions in subsequent searches, thereby reducing the chance of finding new solutions, resulting in the obtained scheduling strategy not being the optimal global solution, and thus failing to achieve the optimal resource allocation and performance of the distribution network. Summary of the invention

[0005] To this end, the technical problem to be solved by the present invention is to overcome the defect that the existing whale algorithm is prone to fall into local optimality during the search stage, resulting in the obtained scheduling strategy not being the optimal global solution, and thus failing to achieve the optimal resource allocation and performance of the distribution network.

[0006] In order to solve the above technical problems, the present invention provides a multi-source microgrid optimization scheduling method, comprising the following steps: Step S1: Obtain mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, construct an objective function with the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal, set power balance constraints, energy storage charge state constraints, power generation constraints, each distributed power source output constraints and ramp rate constraints as constraints, and construct a day-ahead optimization model; Step S2: Obtain the parameter data of the day-ahead scheduling period and input it into the day-ahead optimization model. Solve the day-ahead optimization model through the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy, including: Step S21: Set the number of whales, the maximum number of iterations, the fusion coefficient, the minimum frequency of the sound waves emitted by the bats, and the maximum frequency of the sound waves emitted by the bats; initialize , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. Step S22: Generate a random number in the range of [0, 1] for each whale; Step S23: Determine whether the random number generated by the current whale is less than 0.5; Step S24: if the random number generated by the current whale is greater than or equal to 0.5, the position of the current whale is updated by circling predation; Step S25: If the random number generated by the current whale is less than 0.5, the absolute value of the difference between the fitness of the current whale and the lowest fitness in the historical iteration is used as the judgment coefficient of the current whale; Determine whether the current whale's judgment coefficient is less than 1. If the current whale's judgment coefficient is greater than or equal to 1, use spiral bubble predation to update the current whale's position; if the current whale's judgment coefficient is less than 1, update the current whale's position and speed based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current whale's position and speed. Step S26: Calculate the fitness of each whale based on the current optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Step S27: Determination Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise, return to step S22.

[0007] Preferably, based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bats, the maximum frequency of the sound waves emitted by the bats, and the current position and speed of the whale, the current position and speed of the whale are updated, and the formula is: , , , in, For the The speed of the current whale in this iteration, For the The current whale position of the iteration, For the The current whale position of the iteration, For the The position of the whale with the lowest fitness in the iteration, For the The speed of the current whale in this iteration, is the fusion coefficient, is a random number in the range [0, 1], is the frequency of the bat wave, is the minimum frequency of the sound waves emitted by bats, is the maximum frequency of the sound waves emitted by bats, is a random vector drawn from a regular distribution in the range [0, 1].

[0008] Preferably, the current position of the whale is updated by circling and preying, and the updating formula is: , , in, For the The current whale position of the iteration, is the current iteration number, For the The position of the whale with the lowest fitness in the iteration, is the convergence factor, For the The distance between the current whale and a random individual whale in this iteration, is the swing factor, For the The current whale position.

[0009] Preferably, spiral bubble predation is used to update the current position of the whale, and the update formula is: , , in, For the The current whale position of the iteration, is the current iteration number, is the logarithmic spiral coefficient, is a random number in the range [-1, 1]. For the The distance between the current whale and a random individual whale in this iteration, is a natural constant, For the The position of the whale with the lowest fitness in the iteration, For the The current whale position.

[0010] Preferably, the microgrid distributed power source includes: solar photovoltaic cells, wind turbines, and batteries; The energy conversion device includes: a micro gas turbine, an absorption refrigeration unit, and an electric refrigeration unit.

[0011] Preferably, the step of obtaining the mathematical models of each distributed power source and energy conversion device of the microgrid in the area to be optimized includes: The output model of solar photovoltaic cells is: , in, For the The PV output power of a scheduling cycle is is the power derating factor of the solar panel, is the maximum test power under standard test conditions, is the light intensity under standard test conditions, For solar photovoltaic cells The actual light intensity of the scheduling cycle, is the power temperature coefficient, For solar photovoltaic cells The operating temperature of the scheduling cycle, is the reference temperature of the solar photovoltaic cell; The output model of the wind turbine is: , in, For wind turbines The output power of a scheduling cycle, is the rated power of the wind turbine, is the cut-in wind speed of the wind turbine, is the rated wind speed of the wind turbine, is the cut-out wind speed of the wind turbine, For wind turbines Wind speed during a dispatch cycle; The mathematical model of battery energy storage is: , The mathematical model of battery operating cost is: , in, For the battery The battery charge of a scheduling cycle, For the battery The battery charge of a scheduling cycle, For the battery The output power of a scheduling cycle, is the charging efficiency of the battery, is the discharge efficiency of the battery, is the capacity of the battery, is the self-discharge rate of the battery, For the battery The operating cost of a scheduling cycle, For the battery The state of the scheduling cycle, When it is 1, it means the battery is running. When it is 0, it means the battery is in standby mode. is the maintenance factor of the battery, For the battery Life loss cost of a dispatch cycle; The mathematical model of the power generation cost of a micro gas turbine is: , The mathematical model of the waste heat recovery unit of the micro gas turbine is: , in, is the natural gas consumption, For micro gas turbines The output power of a scheduling cycle, is the power generation efficiency of the micro gas turbine, is the calorific value of natural gas, For micro gas turbines The waste heat power of a dispatch cycle, For micro gas turbines The heating efficiency of a scheduling cycle, is the scheduling interval; The mathematical model of the absorption refrigeration unit is: , in, is the cooling coefficient or heating coefficient of the lithium bromide refrigeration unit, is the cooling capacity or heating capacity of the lithium bromide unit, The heat of the heating source is the waste heat from the exhaust of the combustion engine. The power consumption of the lithium bromide refrigeration unit; The mathematical model of the electric refrigeration unit is: , in, is the refrigeration coefficient of the electric refrigeration unit, is the cooling capacity of the electric refrigeration unit, For electric refrigeration units The power consumption of a scheduling cycle is For the The power purchase and sale price of the power grid in a dispatch cycle, For electric refrigeration units The cooling cost of a scheduling cycle.

[0012] Preferably, the day-ahead optimization model is: , , , , in, is the objective function of the day-ahead optimization model, For the The fuel cost and operation and maintenance cost of the microgrid in a scheduling cycle, For the The pollutant emission treatment cost of the microgrid in a scheduling cycle is: For the The load control compensation cost of the microgrid in a dispatching period is: is the weight of fuel cost and operation and maintenance cost, is the weight of the pollutant emission treatment cost, is the weight of the load control compensation cost, Indicates Distributed power generation The power generation cost of a dispatch cycle is Indicates The operation and maintenance coefficient of the distributed generation, Indicates Distributed power generation The output power of a scheduling cycle, is the scheduling interval, Represents the total number of distributed generation sources in the microgrid, represents the operation and maintenance factor of the micro gas turbine, Indicates the operation and maintenance factor of the absorption chiller unit, Indicates the operation and maintenance factor of the electric refrigeration unit, Indicates that the absorption chiller unit is The output power of a scheduling cycle, Indicates that the electric refrigeration unit is The output power of a scheduling cycle, For the The operating cost of the desulfurization device in a scheduling cycle is For the Scheduling cycle of sewage charges, For the Scheduling cycle of sewage charges, Indicates The unit outage cost of a dispatch cycle is Indicates The load control amount of a dispatching cycle; Power balance constraints: , in, The area to be optimized is The load demand of a dispatching cycle, For the The PV output power of a scheduling cycle is For wind turbines The output power of a scheduling cycle, For the battery The output power of a scheduling cycle, For micro gas turbines The output power of a scheduling cycle, For the The power purchased by the power grid during a dispatch cycle; Output constraints of each distributed power source: , in, For the The output power lower limit of a distributed power source is For the The output power limit of a distributed power source is For the Distributed power generation Output power of each scheduling cycle; Constraints on the ramp rate of each distributed power source: , , in, For the Distributed power generation The ramp rate of a scheduling cycle, For the The ramp rate of a distributed power source is For the Distributed power generation The output power of a scheduling cycle, For the Distributed power generation Output power of each scheduling cycle; Energy storage state of charge constraints: , , in, is the minimum battery charge of the battery, is the maximum battery charge of the battery, For the battery The battery charge of a scheduling cycle, is the maximum charging ratio in a scheduling cycle, is the maximum discharge ratio in a scheduling cycle, For the battery The output power of a scheduling cycle, is the battery charging efficiency, is the battery charging efficiency, is the capacity of the battery; Power generation constraints: , in, For micro gas turbines The output power of a scheduling cycle.

[0013] Preferably, after obtaining the day-ahead optimized scheduling strategy, rolling optimization is used to obtain the target scheduling strategy for each rolling interval within the day, including: In the preset time period before the current rolling interval, the distributed power sources and energy conversion devices of the microgrid in the optimization area are dispatched using the day-ahead optimization dispatching strategy corresponding to the preset time period; Obtain the minimum output power and maximum output power of each distributed power source within the preset time period as the output power lower limit and output power upper limit of the output constraint of each distributed power source in the current rolling interval; Obtain the weighted minimum and weighted maximum values ​​of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid, and the load control compensation cost of the microgrid within the preset time period as the weighted upper limit and weighted lower limit of each cost of the microgrid in the current rolling interval; Obtain the minimum load demand value of the area to be optimized within the preset time period as the lower limit of the load demand of the current rolling interval; Obtain the minimum value and the maximum value of the sum of the cooling capacities of the absorption refrigeration unit and the electric refrigeration unit within the preset time period as the upper limit and the lower limit of the cooling capacity of the current rolling interval; Based on the output power lower limit, output power upper limit, microgrid cost weighted upper limit, microgrid cost weighted lower limit, load demand lower limit, cooling capacity upper limit and cooling capacity lower limit of each distributed power generation output constraint in the current rolling interval, the interval possibility method is used to update the objective function and power balance constraints of the day-ahead optimization model, and set the cold and hot load constraints to obtain the updated objective function and constraints of the day-ahead optimization model; The parameter data of the current rolling interval are obtained and used as the input of the updated day-ahead optimization model. Based on the objective function and constraints of the updated day-ahead optimization model, the objective function of the updated day-ahead optimization model is solved by linear programming to obtain the target scheduling strategy for the current rolling interval.

[0014] Preferably, the interval possibility method is used to update the objective function of the day-ahead optimization model, and the objective function of the updated day-ahead optimization model is: , in, is the objective function of the day-ahead optimization model, is the input variable of the day-ahead optimization model, i.e., the parameter data of the current rolling interval. is the control variable of the day-ahead optimization model, i.e., the output power of each distributed power source and energy conversion device in the current rolling interval in the day-ahead optimization scheduling strategy. is the weighted upper limit of each cost of the microgrid, is the weighted lower limit of each cost of the microgrid, is the probability of weighted intervals of each cost of the microgrid; The power balance constraint of the day-ahead optimization model is updated. The updated power balance constraint is: , in, Indicates the lower limit of load demand, represents the load demand possibility, For wind turbines The corrected output power of the scheduling cycle is: For the The power purchased by the power grid during a dispatch cycle, For the The output power of the battery in a scheduling cycle, Indicates that the area to be optimized is Load demand for a dispatch cycle; The cooling and heating load constraints are: , in, Indicates that the absorption chiller unit is The cooling capacity of a scheduling cycle, Indicates that the electric refrigeration unit is The cooling capacity of a scheduling cycle, Indicates the possible range of cooling capacity, Indicates The upper limit of cooling capacity for a scheduling cycle, Indicates The lower limit of cooling capacity in a scheduling cycle.

[0015] The present invention also provides a multi-source microgrid optimization dispatching system, comprising: The model building module obtains the mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, and constructs the objective function with the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal. The power balance constraint, the state of charge constraint of the energy storage, the power generation constraint, the output constraint of each distributed power source and the ramp rate constraint are set as constraints to build the day-ahead optimization model; The model solving module obtains the parameter data of the day-ahead scheduling period and inputs it into the day-ahead optimization model. The day-ahead optimization model is solved by the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy, including: Initialization unit, used to set the number of whales, maximum number of iterations, fusion coefficient, minimum frequency of sound waves emitted by bats, and maximum frequency of sound waves emitted by bats; initialization , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. A random number generation unit, used to generate a random number in the range of [0, 1] for each whale; A judgment unit, used to judge whether the random number generated by the current whale is less than 0.5; A first position updating unit, configured to update the position of the current whale by circling and preying if the random number generated by the current whale is greater than or equal to 0.5; The second position updating unit is used to take the absolute value of the difference between the fitness of the current whale and the lowest fitness in the historical iteration as the judgment coefficient of the current whale if the random number generated by the current whale is less than 0.5; Determine whether the current whale's judgment coefficient is less than 1. If the current whale's judgment coefficient is greater than or equal to 1, use spiral bubble predation to update the current whale's position; if the current whale's judgment coefficient is less than 1, update the current whale's position and speed based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current whale's position and speed. The fitness calculation unit is used to calculate the fitness of each whale according to the previous day optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Iteration number judgment unit, used to judge Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise the operation of the random number generation unit is executed.

[0016] The above technical solution of the present invention has the following beneficial effects compared with the prior art: The present invention discloses a multi-source microgrid optimization scheduling method and system. The proposed whale-bat algorithm replaces the search and predation phase in the whale algorithm with an improved bat algorithm to update the position of each whale. The bat algorithm draws on the flight and echolocation of bats and uses related variables to simulate the flight process. The flight strategy of the bat algorithm can more flexibly adjust the search direction, helping the whale algorithm to more easily jump out of the local optimal solution, thereby finding a better scheduling strategy. However, the bat algorithm has the problem of over-reliance on local information, which slows down the overall convergence speed and reduces the population diversity, limiting the accuracy and quality of the final solution. Therefore, the improved bat algorithm proposed in the present invention introduces a fusion coefficient and uses the fusion coefficient to control the degree of fusion of speed and position updates. It combines the local search in the bat algorithm with the global search of the whale algorithm to better explore in the search space, so that each whale is not only affected by its own position and speed when updating its position and speed, but also considers the global optimal position and local information. This multi-dimensional information fusion mechanism can more effectively guide whales to search in the solution space and avoid falling into the local optimum. A random number between [0, 1] is also introduced. By adding random disturbances when updating positions, the individual positions can be updated without completely relying on the current solution, thus preventing premature convergence and improving the algorithm's exploration ability on highly complex problems such as multi-source microgrids. It can increase the diversity of the algorithm and help the algorithm escape from the local optimal solution. and random numbers The combination not only maintains the search diversity, but also improves the stability and accuracy of the algorithm in multi-peak and nonlinear optimization problems such as multivariate microgrid scheduling optimization. At the same time, it accelerates the convergence speed of the optimal solution and significantly reduces the calculation time. The obtained optimized scheduling strategy can respond to dynamic changes more quickly while ensuring high accuracy, thereby improving the overall operation efficiency and reliability of the microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein: Figure 1 It is a flow chart of the steps of a multi-source microgrid optimization scheduling method of the present invention.

[0018] Figure 2 This is the architecture diagram of the CCHP microgrid.

[0019] Figure 3 It is the flowchart of the whale-bat algorithm.

[0020] Figure 4 It is a schematic diagram of rolling optimization.

[0021] Figure 5 is the typical daily load curve in summer, Figure 5 (a) is the electric load and renewable energy output curve of the distributed CCHP system. Figure 5 (b) in the figure is the cooling load demand curve of the distributed CCHP system.

[0022] Figure 6 This is a schematic diagram of real-time electricity prices in summer.

[0023] Figure 7 This is a schematic diagram of the day-ahead scheduling optimization results for a typical clear day.

[0024] Figure 8 This is a schematic diagram of the output results of the two types of refrigerators on a typical sunny day.

[0025] Fig. 9 This is a schematic diagram of the optimization results of the day-ahead scheduling on a typical rainy day.

[0026] Fig.10 This is a schematic diagram of the output results of the two types of refrigerators on a typical rainy day.

[0027] Fig.11 It is the day-ahead scheduling cost curve of various scheduling strategies and algorithms on clear days.

[0028] Fig.12 It is the day-ahead scheduling cost curve of various scheduling strategies and algorithms on rainy days.

[0029] Fig.13 It is a schematic diagram of random simulation of prediction error.

[0030] Fig.14 This is a schematic diagram of the real-time dispatch correction results of the microgrid in different time periods under different weather scenarios. Fig.14 (a) is a schematic diagram of the real-time dispatch correction results of the microgrid in different periods under clear weather. Fig.14 (b) is a schematic diagram of the real-time dispatch correction results of the microgrid in different periods under rainy weather.

[0031] Fig.15 This is a schematic diagram of the rolling optimization results of the power load range on sunny days.

[0032] Fig.16 This is a schematic diagram of the rolling optimization results of the cooling load interval on sunny days.

[0033] Fig.17 This is a schematic diagram of the rolling optimization results of the power load interval on rainy days.

[0034] Fig.18 This is a schematic diagram of the rolling optimization results of the cold load interval on rainy days. DETAILED DESCRIPTION

[0035] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0036] Reference Figure 1 As shown, the first embodiment of the present invention provides a multi-source microgrid optimization scheduling method, comprising the following steps: In this embodiment, Figure 2 As shown, Figure 2 This is the architecture diagram of CCHP microgrid. CCHP microgrid can connect various distributed power sources, loads and energy conversion equipment such as refrigerators under the same system, and control and manage them uniformly. It can independently choose to be connected to the grid or off-grid to provide support for cold, hot and electric loads within a certain range, and meet the economical and reliable energy needs of users. Compared with the functional mode of the power grid heating network, CCHP microgrid has demonstrated its significant advantages in renewable energy utilization and integrated energy management by reducing operating costs, realizing energy cascade utilization, improving power quality and reducing environmental pollution.

[0037] Step S1: Obtain mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, take the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal, set power balance constraints, energy storage charge state constraints, power generation constraints, each distributed power source output constraints and ramp rate constraints as constraints, and build a day-ahead optimization model; In this embodiment, preferably, the microgrid distributed power source includes: solar photovoltaic cells, wind turbines, and batteries; The energy conversion device includes: a micro gas turbine, an absorption refrigeration unit, and an electric refrigeration unit.

[0038] Model each distributed power source and energy conversion device of the microgrid in the optimization area to obtain a mathematical model of each distributed power source and energy conversion device; Solar photovoltaic cells are based on the photovoltaic effect and can convert light energy into electrical energy. Photovoltaic power generation does not require fuel, does not produce pollution, is safe and easy to maintain. The output power is mainly determined by conditions such as light intensity, temperature, and duration of illumination.

[0039] In engineering applications, the output model of photovoltaic distributed power generation is: , in, For the The PV output power of a scheduling cycle is is the power derating factor of the solar panel, is the maximum test power under standard test conditions, is the light intensity under standard test conditions, For solar photovoltaic cells The actual light intensity of the scheduling cycle, is the power temperature coefficient, For solar photovoltaic cells The operating temperature of the scheduling cycle, is the reference temperature of the solar photovoltaic cell.

[0040] Wind power generation can convert wind energy into electrical energy. The working principle is mainly that the wind drives the blades on the wind turbine, and finally generates electricity through the action of the speed increaser. There are many types of wind turbines, among which permanent magnet direct-drive wind turbines are more advanced and have the advantages of no need for excitation and better effect at low wind speeds. The proportion of wind turbines in newly installed wind turbines is gradually increasing.

[0041] The power generation capacity of a wind turbine is mainly determined by the cut-out wind speed. and cut-in wind speed Decision. Wind turbines operate at wind speeds greater than It can only start to operate when the wind speed is too fast, but if the wind speed is too fast, it is easy to damage the equipment. In order to prevent accidents, The wind turbine will not start.

[0042] The output model of the wind turbine is: , in, For wind turbines The output power of a scheduling cycle, is the rated power of the wind turbine, is the cut-in wind speed of the wind turbine, is the rated wind speed of the wind turbine, is the cut-out wind speed of the wind turbine, For wind turbines The wind speed of a scheduling cycle.

[0043] Energy Storage (ES) is in the The remaining power in a dispatching cycle is mainly provided by the battery in the The remaining power of the scheduling cycle, The first scheduling cycle The battery charge and discharge capacity in a scheduling cycle and the leakage capacity per hour are used to determine the battery charge and discharge capacity in a scheduling cycle and the leakage capacity per hour.

[0044] The mathematical model of battery energy storage is: , The mathematical model of battery operating cost is: , in, For the battery The battery charge of a scheduling cycle, For the battery The battery charge of a scheduling cycle, For the battery The output power of a scheduling cycle, is the charging efficiency of the battery, is the discharge efficiency of the battery, is the capacity of the battery, is the self-discharge rate of the battery, For the battery The operating cost of a scheduling cycle, For the battery The state of the scheduling cycle, When it is 1, it means the battery is running. When it is 0, it means the battery is in standby mode. is the maintenance factor of the battery, For the battery The life loss cost of a scheduling cycle.

[0045] Micro gas turbines usually lose a lot of excess heat during power generation, so it is difficult to achieve a high level of efficiency. Therefore, the CCHP system's cogeneration mode recycles the excess heat generated during power generation, which can greatly improve the efficiency of the micro gas turbine.

[0046] The mathematical model of the power generation cost of a micro gas turbine is: , The mathematical model of the waste heat recovery unit of the micro gas turbine is: , in, is the natural gas consumption, For micro gas turbines The output power of a scheduling cycle, is the power generation efficiency of the micro gas turbine, is the calorific value of natural gas, For micro gas turbines The waste heat power of a dispatch cycle, For micro gas turbines The heating efficiency of a scheduling cycle, is the scheduling interval; Absorption (lithium bromide) refrigeration units can achieve efficient use of energy, recycle waste heat and low-quality thermal energy, and enhance the economy of the system.

[0047] The mathematical model of the absorption (lithium bromide) refrigeration unit is: , in, is the cooling coefficient or heating coefficient of the lithium bromide refrigeration unit, is the cooling capacity or heating capacity of the lithium bromide unit, The heat of the heating source is the waste heat from the exhaust of the combustion engine. The power consumption of the lithium bromide refrigeration unit; The electric refrigeration unit consumes electricity to drive the compressor to work and cool. Its mathematical model is: , in, is the refrigeration coefficient of the electric refrigeration unit, and in this embodiment, the value is 3. is the cooling capacity of the electric refrigeration unit, For electric refrigeration units The power consumption of a scheduling cycle is For the The power purchase and sale price of the power grid in a dispatch cycle, For electric refrigeration units The cooling cost of a scheduling cycle.

[0048] The optimal operation of a microgrid is a typical planning problem of finding the minimum value of an objective function under multiple objectives and multiple constraints. Considering that the impact of energy consumption on user comfort has a threshold, the traditional scheduling method is improved and a day-ahead scheduling optimization model is established. In this embodiment, under the time division of 24 hours a day, it is approximately assumed that the output of distributed power sources in each time interval is stable and has no fluctuations. On the premise of ensuring the energy consumption and comfort requirements of users, the power generation cost, load compensation cost and environmental cost of various distributed power sources in the microgrid system are quantified, and several sub-goals are weighted and summed according to different demand targets to establish a day-ahead optimization model.

[0049] The day-ahead dispatch optimization model is: , in, is the objective function of the day-ahead optimization model, For the The fuel cost and operation and maintenance cost of the microgrid in a scheduling cycle, For the The pollutant emission treatment cost of the microgrid in a scheduling cycle is: For the The load control compensation cost of the microgrid in a dispatching period is: is the weight of fuel cost and operation and maintenance cost, is the weight of the pollutant emission treatment cost, Weight for compensating cost of load control.

[0050] By expressing all sub-goals as cost present values, a single-objective optimization model is obtained to replace the original multi-objective model. The single-objective optimization model is:

[0051] , in, Indicates Distributed power generation The power generation cost of a dispatch cycle is Indicates The operation and maintenance coefficient of the distributed generation, Indicates Distributed power generation The output power of a scheduling cycle, is the scheduling interval, Represents the total number of distributed generation sources in the microgrid, represents the operation and maintenance factor of the micro gas turbine, Indicates the operation and maintenance factor of the absorption chiller unit, Indicates the operation and maintenance factor of the electric refrigeration unit, Indicates that the absorption chiller unit is The output power of a scheduling cycle, Indicates that the electric refrigeration unit is The output power of a scheduling cycle.

[0052] Micro gas turbines and gas boilers will produce a variety of different air pollutants in the process of burning fuel to generate electricity. The types and quantities of air pollutants produced by different types of distributed power sources are also different. Therefore, emission coefficients are introduced to mark the impact of pollutant emissions from different distributed power sources. These pollutants will have different treatment costs. According to the amount of pollutant emissions from each distributed power source and the unit pollutant treatment cost, the pollution emissions of distributed power sources are used to characterize the environmental friendliness of the system, and converted into present value to obtain the following sub-goals:

[0053] , in, For the The operating cost of the desulfurization device in a scheduling cycle is For the Scheduling cycle of sewage charges, For the Scheduling cycle sewage discharge fee.

[0054] The operation reliability cost of a microgrid is a comprehensive indicator, which reflects all relevant costs required to ensure the normal operation of the microgrid system under various circumstances. Therefore, the two concepts of load control quantity and unit power outage cost are introduced and multiplied to express it, so as to indirectly express the operation reliability cost and the compensation cost of controllable load.

[0055] The formula for the load control compensation cost of the microgrid is: , in, Indicates The unit outage cost of a dispatch cycle is Indicates The load control quantity of a scheduling cycle.

[0056] There are multiple constraints in the optimization model. According to the actual operation of the system, each device has an upper and lower limit on output, and there is a power balance constraint between the device and the load.

[0057] Power balance constraints: , in, The area to be optimized is The load demand of a dispatching cycle, For the The photovoltaic output power of a scheduling cycle, For wind turbines The output power of a scheduling cycle, For the battery The output power of a scheduling cycle, For micro gas turbines The output power of a scheduling cycle, For the The power purchased by the power grid during a dispatch cycle; Output constraints of each distributed power source: , in, For the The output power lower limit of a distributed power source is For the The output power limit of a distributed power source is For the Distributed power generation Output power of each scheduling cycle; Constraints on the ramp rate of each distributed power source: , , in, For the Distributed power generation The ramp rate of a scheduling cycle, For the The ramp rate of a distributed power source is For the Distributed power generation The output power of a scheduling cycle, For the Distributed power generation Output power of each scheduling cycle; Energy storage state of charge constraints: , , in, is the minimum battery charge of the battery, is the maximum battery charge of the battery, For the battery The battery charge of a scheduling cycle, is the maximum charging ratio in a scheduling cycle, is the maximum discharge ratio in a scheduling cycle, For the battery The output power of a scheduling cycle, is the battery charging efficiency, is the battery charging efficiency, is the capacity of the battery; Power generation constraints: , in, For micro gas turbines The output power of a scheduling cycle.

[0058] In the traditional mode, micro gas turbines usually use the "heat / cold power" method to calculate the power generation under the premise of ensuring user comfort. Therefore, in the optimization of the microgrid, the power generation of the micro gas turbine is simplified to a fixed value, which reduces the flexibility of the model. The micro gas turbine cannot adjust the operation of the system, which objectively reduces its positive effect on the economy of the system. Since people's perception of changes in the external environment needs to be based on a certain threshold, the ambient temperature can be flexibly adjusted within this threshold without affecting people's comfort. In this embodiment, the cold and hot loads only determine the basic power generation, which represents the user's comfort guarantee under this scheduling strategy, and the actual power generation can fluctuate by 5% on this basis. This shows that the flexibility of the micro gas turbine can be used to improve the economy of the system on the basis of no or light user perception.

[0059] The bat algorithm draws on the behavior of bats in nature and abstracts it into a swarm intelligence algorithm. The algorithm draws on the flight and echolocation of bats and uses related variables to simulate the flight process. It updates the speed and position of the next generation of bats by adjusting the sound wave frequency and the current global optimal solution to find the optimal solution. The parameter iteration formula of the bat algorithm to simulate the bat echo frequency is: , , , in, For the The frequency of the sound waves emitted by a bat, is the minimum frequency of the sound waves emitted by bats, is the maximum frequency of the sound waves emitted by bats, is a random vector drawn from a regular distribution in the range [0, 1], For the The position of the bat with the lowest fitness in the iteration, For the Bats in The speed of iterations, For the Bats in The speed at the iteration, For the Bats in The position of the iteration, For the Bats in The position of the iteration.

[0060] Each bat in the frequency range The random value is taken inside, the speed of the search agent is compatible with the frequency, and the position of the new solution depends on its new speed.

[0061] The whale algorithm simulates and emulates the hunting methods of humpback whales in the ocean, simulating three hunting mechanisms in the hunting process, in the following order: circling hunting, spiral bubble hunting, and searching hunting.

[0062] (I) Surrounding predation stage The circling hunting phase in the whale algorithm simulates the coordinated hunting behavior of a group of whales in the search space. In this phase, the whale adjusts its position to find the global optimal solution, and dynamically updates its position during the iteration process to approach the optimal solution. The position update formula is: , , in, For the The current whale position of the iteration, is the current iteration number, For the The position of the whale with the lowest fitness in the iteration, is the convergence factor, For the The distance between the current whale and a random individual whale in this iteration, is the swing factor, For the The current whale position.

[0063] (II) Spiral bubble predation stage The spiral bubble hunting phase in the whale algorithm is to guide the whales to move in the search space by generating spiral bubbles. Each whale adjusts its position according to the generated spiral bubbles in the hope of better approximating the optimal solution. In this phase, the spiral bubbles are applied iteratively multiple times to help the whales gradually optimize their positions during the search process. The whales update their positions according to the following formula: , , in, For the The current whale position of the iteration, is the current iteration number, is the logarithmic spiral coefficient, is a random number in the range [-1, 1]. For the The distance between the current whale and a random individual whale in this iteration, is a natural constant, For the The position of the whale with the lowest fitness in the iteration, For the The current whale position.

[0064] 3. Search and predation phase The search and predation phase of the whale algorithm simulates the process of whales searching for potential solutions in the search space. Each whale changes its position and uses the search strategy to find the target position, thereby gradually optimizing its search direction. This phase uses multiple iterations to improve the efficiency of the algorithm to find the global optimal solution. The formula for the search and predation phase is: , , The global search capability of the whale algorithm is excellent, but its search phase updates individual positions by randomly selecting target positions. In a complex optimization space, randomly selecting target positions may cause the algorithm to repeatedly search near a local optimal solution and it is difficult to jump out of this range. Therefore, this method is prone to fall into local optimality in complex, multi-peak optimization problems, especially in scheduling problems. As the iteration proceeds, the algorithm is prone to lose diversity, which makes it easier for the algorithm to fall into similar solutions in subsequent searches, thereby reducing the chance of finding new solutions, resulting in the obtained scheduling strategy not being the optimal global solution, and thus failing to achieve the optimal resource allocation and performance of the distribution network. In contrast, the bat algorithm draws on the flight and echolocation of bats and uses related variables to simulate the flight process. The flight strategy of the bat algorithm can more flexibly adjust the search direction, helping the whale algorithm to more easily jump out of the local optimal solution and find a better scheduling strategy. However, the bat algorithm lacks memory of local optimal solutions during the optimization process. This problem mainly stems from its algorithm mechanism. According to the design of the bat algorithm, the behavior of each bat simulates the search mode of acoustic echolocation in nature, but the algorithm does not have a special structure to retain and utilize the local optimal solution obtained in each iteration. In other words, the bat algorithm only updates the current position and speed in each iteration, and does not record the historical optimal solution. This lack of memory structure causes the algorithm to easily deviate from the global optimal search path when searching for potential local optimal solutions. Therefore, when the bat algorithm is used directly to solve the multi-microgrid scheduling strategy, it will cause delays in the multi-microgrid scheduling strategy, which is particularly obvious in the case of dynamic load changes or demand fluctuations. It cannot respond to external changes in time, reducing the stability and reliability of the microgrid. In addition, the lack of a global perspective search method will lead to unsatisfactory optimization results, making the obtained multi-microgrid scheduling strategy not a global optimal solution, and unable to complete the reasonable configuration of multi-source microgrid resources.

[0065] In order to give full play to the local positioning advantage of the bat algorithm and the global search advantage of the whale algorithm, the present invention combines the two algorithms to solve the shortcomings of the two algorithms in their respective disadvantageous stages, reduce the number of algorithm iterations, make the convergence effect more obvious, and improve the search speed and accuracy of the algorithm.

[0066] like Figure 3 As shown, Figure 3 This is the flow chart of the whale-bat algorithm. Step S2: Obtain the parameter data of the day-ahead scheduling period and input it into the day-ahead optimization model. The day-ahead optimization model is solved by the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy.

[0067] Step S21: Set the number of whales, the maximum number of iterations, the fusion coefficient, the minimum frequency of the sound waves emitted by the bats, and the maximum frequency of the sound waves emitted by the bats; initialize , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. Step S22: Generate a random number P in the range of [0, 1] for each whale; Step S23: Determine whether the random number P generated by the current whale is less than 0.5; Step S24: if the random number P generated by the current whale is greater than or equal to 0.5, the position of the current whale is updated by circling predation; Step S25: If the random number P generated by the current whale is less than 0.5, the absolute value of the difference between the current whale's fitness and the lowest fitness in the historical iteration is calculated. , as the judgment coefficient of the current whale; Determine the current whale's judgment coefficient Is it less than 1? If the current whale's judgment coefficient If the current whale's judgment coefficient is greater than or equal to 1, the spiral bubble predation is used to update the current whale's position; if the current whale's judgment coefficient is If it is less than 1, the current position and speed of the whale are updated based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current position and speed of the whale; Step S26: Calculate the fitness of each whale based on the current optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Step S27: Determination Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise, return to step S22.

[0068] The fitness of each whale is calculated according to the day-ahead optimization model and the position of each whale, and the fitness of each whale represents an objective function value of the day-ahead optimization model.

[0069] The parameter data of the day-ahead scheduling period include: Power derating factor of solar panels, maximum test power under standard test conditions, light intensity under standard test conditions, reference temperature of solar photovoltaic cells, rated power of wind turbines, cut-in wind speed of wind turbines, rated wind speed of wind turbines, cut-out wind speed of wind turbines, charging efficiency of batteries, discharge efficiency of batteries, capacity of batteries, self-discharge rate of batteries, maintenance factor of batteries, power generation efficiency of micro gas turbines, calorific value of natural gas, cooling coefficient or heating coefficient of lithium bromide refrigeration unit, cooling capacity or heating capacity of lithium bromide unit, heat of heating source, cooling capacity of electric refrigeration unit, actual light intensity of solar photovoltaic cells in the day-ahead scheduling cycle, solar photovoltaic cells in each scheduling cycle in the day-ahead scheduling cycle The operating temperature of the wind turbine, the wind speed of the wind turbine in each dispatching cycle in the day-ahead dispatching cycle, the battery charge of the battery in each dispatching cycle in the day-ahead dispatching cycle, the operating cost of the battery in each dispatching cycle in the day-ahead dispatching cycle, the life loss cost of the battery in each dispatching cycle in the day-ahead dispatching cycle, the waste heat power of the micro gas turbine in each dispatching cycle in the day-ahead dispatching cycle, the heating efficiency of the micro gas turbine in each dispatching cycle in the day-ahead dispatching cycle, the power consumption of the electric refrigeration unit in each dispatching cycle in the day-ahead dispatching cycle, the grid purchase and sale price in each dispatching cycle in the day-ahead dispatching cycle, the cooling cost of the electric refrigeration unit in each dispatching cycle in the day-ahead dispatching cycle, and the load demand of the area to be optimized in each dispatching cycle in the day-ahead dispatching cycle.

[0070] In this embodiment, preferably, the improved bat algorithm is used to update the current position of the whale, and the update formula is:

[0071] , , , in, For the The speed of the current whale in this iteration, For the The current whale position of the iteration, For the The current whale position of the iteration, For the The position of the whale with the lowest fitness in the iteration, For the The speed of the current whale in this iteration, is the fusion coefficient, is a random number in the range [0, 1], is the frequency of the bat wave, is the minimum frequency of the sound waves emitted by bats, is the maximum frequency of the sound waves emitted by bats, is a random vector drawn from a regular distribution in the range [0, 1].

[0072] The present invention solves the shortcomings of the whale and bat algorithms in their respective disadvantageous stages, adopts the bat algorithm to help the whale algorithm jump out of the local optimal solution, embeds the bat algorithm part into the whale algorithm search stage, and introduces the fusion coefficient and a random number in the range [0, 1] , overcome the problem of the bat algorithm's over-reliance on local information, use the fusion coefficient to control the degree of fusion of speed and position updates, combine the local search advantage of the bat algorithm with the global search capability of the whale algorithm, and more effectively avoid local optimal traps, so as to better explore in the search space, so that each whale is not only affected by its own position and speed when updating its position and speed, but also considers the global optimal position and local information. This multi-dimensional information fusion mechanism can more effectively guide whales to search in the solution space and avoid falling into local optimality; it also introduces a random number in the range [0, 1] By adding random disturbances when updating positions, the individual positions can be updated without completely relying on the current solution, thus preventing premature convergence and improving the algorithm's exploration ability on highly complex problems such as multi-source microgrids. It can increase the diversity of the algorithm and help the algorithm escape from the local optimal solution. and random numbers The combination not only maintains the search diversity, but also improves the stability and accuracy of the algorithm in multi-peak and nonlinear optimization problems such as multivariate microgrid scheduling optimization. At the same time, it accelerates the convergence speed of the optimal solution and significantly reduces the calculation time. The obtained optimized scheduling strategy can respond to dynamic changes more quickly while ensuring high accuracy, thereby improving the overall operation efficiency and reliability of the microgrid.

[0073] like Figure 4 As shown, Figure 4 It is a rolling optimization diagram. The present invention adopts a rolling optimization strategy that takes into account the unified prediction error to optimize the second time scale, namely real-time scheduling. The scheduling value optimized the day before and the actual value of the previous time period are used as the reference benchmark value for real-time scheduling. A second round of rolling prediction optimization is performed every 15 minutes for the subsequent 4 hours to optimize the energy matching between each distributed power source, electric load, and cold and hot load.

[0074] In multi-source microgrid systems, load demand and renewable energy output (such as photovoltaics, wind power, etc.) are often uncertain. The traditional rolling interval method mainly relies on single-point prediction, which only adjusts the output according to the updated data in each rolling interval without explicitly considering the range impact of these uncertainties. Through the interval possibility method, the uncertainty range can be directly introduced into the objective function and constraints of the optimization model, so that the model can better cope with fluctuations in the actual environment. This robustness is more suitable for renewable energy power generation scenarios in multi-source microgrids than traditional rolling optimization, because this form of power generation has large volatility, and the interval possibility method helps to ensure the effectiveness of the model under fluctuations; and in traditional rolling optimization, as load fluctuations increase, the model often needs to be updated and adjusted more frequently, while the interval possibility method only needs to define a probability range, which can reduce the frequency of rolling adjustments, thereby reducing computational complexity and control frequency.

[0075] The interval possibility method allows the use of probabilistic expressions for constraints, so that the model not only satisfies the constraints under expected circumstances, but also satisfies them under a specific probability; The interval possibility method allows the use of probabilistic expressions for constraints, so that the model not only satisfies the constraints under expected circumstances, but also satisfies them under a specific probability, which helps to balance reliability and economy in system operation.

[0076] In this embodiment, preferably, after obtaining the day-ahead optimized scheduling strategy, rolling optimization is used to obtain the target scheduling strategy for each rolling interval within the day, including: In the preset time period before the current rolling interval, the distributed power sources and energy conversion devices of the microgrid in the optimization area are dispatched using the day-ahead optimization dispatching strategy corresponding to the preset time period; Obtain the minimum output power and maximum output power of each distributed power source within the preset time period as the output power lower limit and output power upper limit of the output constraint of each distributed power source in the current rolling interval; Obtain the weighted minimum and maximum values ​​of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid, and the load control compensation cost of the microgrid within the preset time period as the weighted upper limit and the weighted lower limit of each cost of the microgrid in the current rolling interval; Obtain the minimum load demand value of the area to be optimized within the preset time period as the lower limit of the load demand of the current rolling interval; Obtain the minimum value and the maximum value of the sum of the cooling capacities of the absorption refrigeration unit and the electric refrigeration unit within the preset time period as the upper limit and the lower limit of the cooling capacity of the current rolling interval; Based on the output power lower limit, output power upper limit, microgrid cost weighted upper limit, microgrid cost weighted lower limit, load demand lower limit, cooling capacity upper limit and cooling capacity lower limit of each distributed power generation output constraint in the current rolling interval, the interval possibility method is used to update the objective function and power balance constraints of the day-ahead optimization model, and set the cold and hot load constraints to obtain the updated objective function and constraints of the day-ahead optimization model; The parameter data of the current rolling interval are obtained and used as the input of the updated day-ahead optimization model. Based on the objective function and constraints of the updated day-ahead optimization model, the objective function of the updated day-ahead optimization model is solved by linear programming to obtain the target scheduling strategy for the current rolling interval.

[0077] Different from the interior point method and other methods, the present invention adopts the interval possibility method for processing. The interval possibility is usually used to characterize the possibility that a certain interval is greater than or equal to another interval.

[0078] , , in, is the weighted lower limit of each cost of the microgrid, is the weighted upper limit of each cost of the microgrid, is the weighted interval possibility of each cost of the microgrid, The input variables of the day-ahead optimization model are the parameter data of the current rolling interval, including external conditions and control parameters, such as wind speed, solar radiation intensity, load demand and other random variables in the current scheduling cycle. is the control variable of the day-ahead optimization model, i.e., the output power of each distributed power source and energy conversion device in the current rolling interval in the day-ahead optimization scheduling strategy. This is the objective function of the day-ahead scheduling optimization model.

[0079] The processing of the interval possibility method is as follows. Let and Is the interval The number within, and ,when Degenerates into When the interval is possible for: , Similarly, if Degenerates into When the interval is possible for: , Furthermore, we can get the conclusion of the interval possibility method: , , The rolling optimization strategy in the second time scale can perform a deterministic transformation of the power balance constraint of the day-ahead optimization model according to the interval possibility method, taking into account the unified prediction error: , , , , in, The area to be optimized is The load demand of a dispatching cycle, The area to be optimized is The lower limit of the load in a scheduling cycle, The area to be optimized is The load limit of a scheduling cycle, For the The forward photovoltaic output power of a scheduling cycle is For the The negative photovoltaic output power of a scheduling cycle, For micro gas turbines The forward output power of a scheduling cycle is For micro gas turbines The negative output power of a scheduling cycle, is the lower limit of residual load, The upper limit of the residual load.

[0080] Therefore, the power balance constraint of the day-ahead optimization model can be updated. The updated power balance constraint is:

[0081] in, Indicates the lower limit of load demand, represents the load demand possibility, For wind turbines The corrected output power of the scheduling cycle is: For the The power purchased by the power grid during a dispatch cycle, For the The output power of the battery in a scheduling cycle, Indicates that the area to be optimized is The load demand of a scheduling cycle.

[0082] The cooling and heating load constraints are: , in, Indicates that the absorption chiller unit is The cooling capacity of a scheduling cycle, Indicates that the electric refrigeration unit is The cooling capacity of a scheduling cycle, Indicates the possible range of cooling capacity, Indicates The upper limit of cooling capacity for a scheduling cycle, Indicates The lower limit of cooling capacity in a scheduling cycle.

[0083] The remaining constraints remain unchanged, so that the objective function and constraints of the updated day-ahead optimization model are linear, so a linear programming solution can be performed for this optimization.

[0084] The introduction of the interval possibility method into the rolling optimization allows the model to enhance its adaptability to uncertainty while satisfying the advantages of rolling adjustment, and ensures that the system's needs and constraints are met within a given probability range, highlighting the unique advantages of this method in dealing with dynamic loads and fluctuations in multi-source power generation systems. Compared with the traditional method that only uses rolling interval optimization, the interval possibility method has significant improvements in robustness and flexibility, thereby improving the overall dispatch performance of the multi-source microgrid system.

[0085] Since practical control is almost always affected by external interference and various uncontrollable factors, or large errors in prediction will lead to inaccurate optimization results and make it difficult to apply in actual scenarios. Therefore, the feedback correction link in rolling optimization is of great significance. Through its feedback information, a closed loop is formed to enhance the robustness of the entire link and the anti-interference ability of the system, which plays a very important role in reducing prediction errors and improving system availability. The result of feedback correction will become a reference for constraints. Because in short-term scheduling, if the output power of each device is calculated from zero in each rolling interval, it may lead to large instantaneous adjustments and unstable load changes. The role of feedback correction is to accumulate the error information of the previous period, balance the output power requirements of each period, and avoid extreme load fluctuations in the rolling interval.

[0086] In this embodiment, different fusion coefficients are set in different time periods. It is possible to balance feedback information and prediction information in the day-ahead optimization model. For example, the fusion coefficients of the whale-bat algorithm are set to 0.5, 0.3, and 0.7, respectively, and the day-ahead optimization scheduling model of a small microgrid system containing distributed equipment such as photovoltaics, wind power generation, micro gas turbines, and battery energy storage is solved.

[0087] The load demand prediction value, actual feedback load value, and load demand correction value of the small microgrid system are shown in Table 1.

[0088] Table 1 Scheduling period (t) Load demand forecast (kW) Actual feedback load (kW) Fusion coefficient α Load demand correction (kW) 0:00-1:00 120 115 0.5 117.5 1:00-2:00 130 125 0.3 126.5 2:00-3:00 140 135 0.7 138.5 3:00-4:00 150 145 0.5 147.5 4:00-5:00 160 158 0.5 159 If the fusion coefficient When , it can balance the load demand forecast value and the actual feedback load value, and can better track the actual load demand, which is suitable for scenarios with stable load changes; when the fusion coefficient When the load demand correction value is more inclined to the actual feedback load value, it is suitable for situations where the load fluctuation is large and can adjust the power output more quickly; when the fusion coefficient When the load demand correction value is closer to the load demand forecast value, it is suitable for scenarios with small load fluctuations to avoid frequent adjustments caused by short-term fluctuations.

[0089] The second embodiment simulates the typical daily load demand of an office building, and optimizes and calculates the operation of the distributed CCHP system through a multi-element microgrid optimization scheduling method proposed by the present invention.

[0090] like Figure 5 As shown, Figure 5 is the typical daily load curve in summer, Figure 5 (a) is the electric load and renewable energy output curve of the distributed CCHP system. Figure 5 (b) in the figure is the cooling load demand curve of the distributed CCHP system.

[0091] like Figure 6 As shown, Figure 6 This is a diagram of real-time electricity prices in summer. In order to improve the utilization efficiency of power resources and reduce waste, the office building uses real-time electricity prices to implement charges.

[0092] The basic parameters of each device in the microgrid are shown in Table 2.

[0093] Table 2 Device Type Photovoltaic Fan gas turbine Battery Controllable load Service life (years) 20 15 10 15 - Installation cost (10,000 yuan / kW) 6.650 2.385 1.220 3.356 - Maintenance factor (yuan / kW) 0.0096 0.0296 0.0401 0.08687 - Power lower limit (kW) 0 0 60 -50 -20 Power limit (kW) 40 60 200 80 20

[0094] The pollutant emission parameters of the power generation unit are shown in Table 3.

[0095] Table 3 Table 3

[0096] In addition, the basic parameters of the battery are: charging and discharging efficiency is 95%, the minimum charge is 40kWh, the total capacity is 200kWh, and the initial charge state is at the minimum value.

[0097] This embodiment designs two different scenarios for the operation of the grid-connected AC microgrid. Considering that wind and photovoltaic power generation are greatly affected by weather, two scenarios of sunny day and rainy day are designed, and the whale-bat algorithm designed by the present invention is used to solve the day-ahead scheduling model.

[0098] The optimal results of the day-ahead dispatch for a typical sunny day are as follows: Figure 7 , 8 As shown, Figure 7 This is a schematic diagram of the optimization results of the day-ahead scheduling on a typical sunny day. Figure 8 This is a schematic diagram of the output results of the two types of refrigerators on a typical sunny day.

[0099] On rainy days, the output of each distributed power source and the output of the refrigerator are as follows: Fig. 9 , Fig.10 As shown, Fig. 9 This is a schematic diagram of the optimization results of the day-ahead scheduling on a typical rainy day. Fig.10 This is a schematic diagram of the output results of the two types of refrigerators on a typical rainy day.

[0100] like Fig.11 , 12 As shown, Fig.11 is the day-ahead scheduling cost curve for various scheduling strategies and algorithms on a clear day, Fig.12 The day-ahead dispatch cost curves of various dispatch strategies and algorithms for rainy days are shown in Figure 2. By comparison, it can be found that the trends of operating costs and load changes in each period are roughly similar, showing a bimodal trend.

[0101] The examples under different weather conditions are simulated to obtain the cost comparison of various commonly used algorithms. The results after running are shown in Table 4.

[0102] Table 4

[0103] It can be clearly seen from Table 4 that the day-ahead dispatching cost obtained by the whale-bat algorithm proposed in the present invention in both sunny and rainy day scenarios is lower than that of other intelligent algorithms, which shows that the adaptability and robustness of the whale-bat algorithm under different weather conditions are better than other intelligent optimization methods. The whale-bat algorithm proposed in the present invention not only improves the search efficiency, but also reduces the dispatching cost fluctuation caused by weather changes, and is more flexible in resource allocation in microgrids.

[0104] The main purpose of real-time optimization scheduling is to reduce the adverse effects caused by the error between the data prediction and the actual data, which is mainly achieved by real-time regulation of the joint rapid response of the battery and the controllable load. The present invention uses a unified prediction error percentage to represent the prediction error, and the error amount is extracted by Monte Carlo random sampling. The actual prediction error amount sampling simulation of sunny days and rainy days is as follows Fig.13 As shown, Fig.13 Schematic diagram of random simulation of prediction error.

[0105] Fig.14 This is a schematic diagram of the real-time dispatch correction results of the microgrid in different time periods under different weather scenarios. Fig.14 (a) is a schematic diagram of the real-time dispatch correction results of the microgrid in different periods under clear weather. Fig.14 (b) in the figure is a schematic diagram of the real-time dispatch correction results of the microgrid in different periods under rainy weather. The fluctuations of the battery and the controllable load are simulated respectively, which shows the tracking effect of the actual load change based on the optimization results of the first time scale in the second time scale of the microgrid, which is mainly achieved by adjusting the access and output ratio of the controllable load and the battery. Among them, the adjustment amount of the battery represents the discharge output with a positive value, and the charging consumption with a negative value. The positive value of the controllable load adjustment represents an increase in load, and the negative value represents a decrease in load.

[0106] After the system's day-ahead scheduling optimization, real-time rolling optimization is performed. If the output exceeds the forecast interval, the excess energy is not processed and is constrained to the upper or lower limit of the interval. The simulation results of the optimal scheduling of electric load and cooling load under clear weather are obtained, such as Fig.15 , 16 As shown, Fig.15 This is a schematic diagram of the rolling optimization results of the power load interval on a sunny day. Fig.16 This is a schematic diagram of the rolling optimization results of the cooling load interval on a sunny day. Fig.17 , 18 As shown, Fig.17 This is a schematic diagram of the rolling optimization results of the power load interval on rainy days. Fig.18 This is a schematic diagram of the rolling optimization results of the cold load interval on rainy days.

[0107] The present invention comprehensively considers economy and environmental protection, flexibly configures various demand weights, and uniformly configures and manages the power generation cost and environmental cost of the multi-source system; for the CCHP system, an improved scheduling strategy is proposed, which considers the thermal-electric coupling to configure the basic electric power adjustment range of micro-sources such as micro-gas turbines, and pursues the economic operation of the system under the premise of ensuring load demand, system power balance and user comfort.

[0108] When simulating the operation of a multi-source microgrid, the results of the day-ahead dispatch are referred to, and the rolling optimization based on the uncertainty of the prediction interval is used as the optimal dispatching strategy at the second time scale to track the changes of each micro-source and load. The optimized dispatching strategy can improve the overall economy and environmental benefits of the system under different typical weather conditions.

[0109] The third embodiment also provides a multi-source microgrid optimization dispatching system, including: The model building module obtains the mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, and constructs the objective function with the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal. The power balance constraint, the state of charge constraint of the energy storage, the power generation constraint, the output constraint of each distributed power source and the ramp rate constraint are set as constraints to build the day-ahead optimization model; The model solving module obtains the parameter data of the day-ahead scheduling period and inputs it into the day-ahead optimization model. The day-ahead optimization model is solved by the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy, including: Initialization unit, used to set the number of whales, maximum number of iterations, fusion coefficient, minimum frequency of sound waves emitted by bats, and maximum frequency of sound waves emitted by bats; initialization , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. A random number generation unit, used to generate a random number in the range of [0, 1] for each whale; A judgment unit, used to judge whether the random number generated by the current whale is less than 0.5; A first position updating unit, configured to update the position of the current whale by circling and preying if the random number generated by the current whale is greater than or equal to 0.5; The second position updating unit is used to take the absolute value of the difference between the fitness of the current whale and the lowest fitness in the historical iteration as the judgment coefficient of the current whale if the random number generated by the current whale is less than 0.5; Determine whether the current whale's judgment coefficient is less than 1. If the current whale's judgment coefficient is greater than or equal to 1, use spiral bubble predation to update the current whale's position; if the current whale's judgment coefficient is less than 1, update the current whale's position and speed based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current whale's position and speed. The fitness calculation unit is used to calculate the fitness of each whale according to the previous day optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Iteration number judgment unit, used to judge Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise the operation of the random number generation unit is executed.

[0110] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0111] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0112] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0113] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0114] Obviously, the above embodiments are merely examples for the purpose of clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the present invention.

Claims

1. A multi-source microgrid optimization scheduling method, characterized in that: The following steps are involved: Step S1: Obtain mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, construct an objective function with the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal, set power balance constraints, energy storage charge state constraints, power generation constraints, each distributed power source output constraints and ramp rate constraints as constraints, and construct a day-ahead optimization model; Step S2: Obtain the parameter data of the day-ahead scheduling period and input it into the day-ahead optimization model. Solve the day-ahead optimization model through the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy, including: Step S21: Set the number of whales, the maximum number of iterations, the fusion coefficient, the minimum frequency of the sound waves emitted by the bats, and the maximum frequency of the sound waves emitted by the bats; initialize , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. Step S22: Generate a random number in the range of [0, 1] for each whale; Step S23: Determine whether the random number generated by the current whale is less than 0.5; Step S24: if the random number generated by the current whale is greater than or equal to 0.5, the position of the current whale is updated by circling predation; Step S25: If the random number generated by the current whale is less than 0.5, the absolute value of the difference between the fitness of the current whale and the lowest fitness in the historical iteration is used as the judgment coefficient of the current whale; Determine whether the current whale's judgment coefficient is less than 1. If the current whale's judgment coefficient is greater than or equal to 1, use spiral bubble predation to update the current whale's position; if the current whale's judgment coefficient is less than 1, update the current whale's position and speed based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current whale's position and speed. Step S26: Calculate the fitness of each whale based on the current optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Step S27: Determination Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise, return to step S22.

2. A multi-source microgrid optimization scheduling method according to claim 1, characterized in that: Based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current position and speed of the whale, the current position and speed of the whale are updated. The formula is: , ， ， in, For the The speed of the current whale in this iteration, For the The current whale position of the iteration, For the The current whale position of the iteration, For the The position of the whale with the lowest fitness in the iteration, For the The speed of the current whale in this iteration, is the fusion coefficient, is a random number in the range [0, 1], is the frequency of the bat wave, is the minimum frequency of the sound waves emitted by bats, is the maximum frequency of the sound waves emitted by bats, is a random vector drawn from a regular distribution in the range [0, 1].

3. A multi-source microgrid optimization scheduling method according to claim 1, characterized in that: Update the current position of the whale by circling and hunting. The update formula is: ， ， in, For the The current whale position of the iteration, is the current iteration number, For the The position of the whale with the lowest fitness in the iteration, is the convergence factor, For the The distance between the current whale and a random individual whale in this iteration, is the swing factor, For the The current whale position.

4. A multi-source microgrid optimization scheduling method according to claim 1, characterized in that: The spiral bubble predation is used to update the current position of the whale. The update formula is: ， ， in, For the The current whale position of the iteration, is the current iteration number, is the logarithmic spiral coefficient, is a random number in the range [-1, 1]. For the The distance between the current whale and a random individual whale in this iteration, is a natural constant, For the The position of the whale with the lowest fitness in the iteration, For the The current whale position.

5. A multi-source microgrid optimization scheduling method according to claim 1, characterized in that: Microgrid distributed power sources include: solar photovoltaic cells, wind turbines, and batteries; The energy conversion device includes: a micro gas turbine, an absorption refrigeration unit, and an electric refrigeration unit.

6. A multi-source microgrid optimization scheduling method according to claim 5, characterized in that: The method of obtaining the mathematical model of each distributed power source and energy conversion device of the microgrid in the area to be optimized includes: The output model of solar photovoltaic cells is: ， in, For the The PV output power of a scheduling cycle is is the power derating factor of the solar panel, is the maximum test power under standard test conditions, is the light intensity under standard test conditions, For solar photovoltaic cells The actual light intensity of the scheduling cycle, is the power temperature coefficient, For solar photovoltaic cells The operating temperature of the scheduling cycle, is the reference temperature of the solar photovoltaic cell; The output model of the wind turbine is: ， in, For wind turbines The output power of a scheduling cycle, is the rated power of the wind turbine, is the cut-in wind speed of the wind turbine, is the rated wind speed of the wind turbine, is the cut-out wind speed of the wind turbine, For wind turbines Wind speed during a dispatch cycle; The mathematical model of battery energy storage is: ， The mathematical model of battery operating cost is: ， in, For the battery The battery charge of a scheduling cycle, For the battery The battery charge of a scheduling cycle, For the battery The output power of a scheduling cycle, is the charging efficiency of the battery, is the discharge efficiency of the battery, is the capacity of the battery, is the self-discharge rate of the battery, For the battery The operating cost of a scheduling cycle, For the battery The state of the scheduling cycle, When it is 1, it means the battery is running. When it is 0, it means the battery is in standby mode. is the maintenance factor of the battery, For the battery Life loss cost of a dispatch cycle; The mathematical model of the power generation cost of a micro gas turbine is: ， The mathematical model of the waste heat recovery unit of the micro gas turbine is: ， in, is the natural gas consumption, For micro gas turbines The output power of a scheduling cycle, is the power generation efficiency of the micro gas turbine, is the calorific value of natural gas, For micro gas turbines The waste heat power of a dispatch cycle, For micro gas turbines The heating efficiency of a scheduling cycle, is the scheduling interval; The mathematical model of the absorption refrigeration unit is: ， in, is the cooling coefficient or heating coefficient of the lithium bromide refrigeration unit, is the cooling capacity or heating capacity of the lithium bromide unit, The heat of the heating source is the waste heat from the exhaust of the combustion engine. The power consumption of the lithium bromide refrigeration unit; The mathematical model of the electric refrigeration unit is: ， in, is the refrigeration coefficient of the electric refrigeration unit, is the cooling capacity of the electric refrigeration unit, For electric refrigeration units The power consumption of a scheduling cycle is For the The power purchase and sale price of the power grid in a dispatch cycle, For electric refrigeration units The cooling cost of a scheduling cycle.

7. A multi-source microgrid optimization scheduling method according to claim 6, characterized in that: The day-ahead optimization model is: ， ， ， ， in, is the objective function of the day-ahead optimization model, For the The fuel cost and operation and maintenance cost of the microgrid in a scheduling cycle, For the The pollutant emission treatment cost of the microgrid in a scheduling cycle is: For the The load control compensation cost of the microgrid in a dispatching period is: are the weights of fuel cost and operation and maintenance cost, is the weight of the pollutant emission treatment cost, is the weight of the load control compensation cost, Indicates Distributed power generation The power generation cost of a dispatch cycle is Indicates The operation and maintenance coefficient of the distributed generation, Indicates Distributed power generation The output power of a scheduling cycle, is the scheduling interval, Represents the total number of distributed generation sources in the microgrid, represents the operation and maintenance factor of the micro gas turbine, Indicates the operation and maintenance factor of the absorption chiller unit, Indicates the operation and maintenance factor of the electric refrigeration unit, Indicates that the absorption chiller unit is The output power of a scheduling cycle, Indicates that the electric refrigeration unit is The output power of a scheduling cycle, For the The operating cost of the desulfurization device in a scheduling cycle is For the Scheduling cycle of sewage charges, For the Scheduling cycle of sewage charges, Indicates The unit outage cost of a dispatch cycle is Indicates The load control amount of a dispatching cycle; Power balance constraints: ， in, The area to be optimized is The load demand of a dispatching cycle, For the The photovoltaic output power of a scheduling cycle, For wind turbines The output power of a scheduling cycle, For the battery The output power of a scheduling cycle, For micro gas turbines The output power of a scheduling cycle, For the The power purchased by the power grid during a dispatch cycle; Output constraints of each distributed power source: ， in, For the The output power lower limit of a distributed power source is For the The output power limit of a distributed power source is For the Distributed power generation Output power of each scheduling cycle; Constraints on the ramp rate of each distributed power source: ， ， in, For the Distributed power generation The ramp rate of a scheduling cycle, For the The ramp rate of a distributed power source is For the Distributed power generation The output power of a scheduling cycle, For the Distributed power generation Output power of each scheduling cycle; Energy storage state of charge constraints: ， ， in, is the minimum battery charge of the battery, is the maximum battery charge of the battery, For the battery The battery charge of a scheduling cycle, is the maximum charging ratio in a scheduling cycle, is the maximum discharge ratio in a scheduling cycle, For the battery The output power of a scheduling cycle, is the battery charging efficiency, is the battery charging efficiency, is the capacity of the battery; Power generation constraints: ， in, For micro gas turbines The output power of a scheduling cycle.

8. A multi-source microgrid optimization scheduling method according to claim 7, characterized in that: After obtaining the day-ahead optimized scheduling strategy, rolling optimization is used to obtain the target scheduling strategy for each rolling interval within the day, including: In the preset time period before the current rolling interval, the distributed power sources and energy conversion devices of the microgrid in the optimization area are dispatched using the day-ahead optimization dispatching strategy corresponding to the preset time period; Obtain the minimum output power and maximum output power of each distributed power source within the preset time period as the output power lower limit and output power upper limit of the output constraint of each distributed power source in the current rolling interval; Obtain the weighted minimum and maximum values ​​of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid, and the load control compensation cost of the microgrid within the preset time period as the weighted upper limit and the weighted lower limit of each cost of the microgrid in the current rolling interval; Obtain the minimum load demand value of the area to be optimized within the preset time period as the lower limit of the load demand of the current rolling interval; Obtain the minimum value and the maximum value of the sum of the cooling capacities of the absorption refrigeration unit and the electric refrigeration unit within the preset time period as the upper limit and the lower limit of the cooling capacity of the current rolling interval; Based on the output power lower limit, output power upper limit, microgrid cost weighted upper limit, microgrid cost weighted lower limit, load demand lower limit, cooling capacity upper limit and cooling capacity lower limit of each distributed power generation output constraint in the current rolling interval, the interval possibility method is used to update the objective function and power balance constraints of the day-ahead optimization model, and set the cold and hot load constraints to obtain the updated objective function and constraints of the day-ahead optimization model; The parameter data of the preset time period in the current rolling interval are obtained and used as the input of the updated day-ahead optimization model. Based on the objective function and constraints of the updated day-ahead optimization model, the objective function of the updated day-ahead optimization model is solved by linear programming to obtain the target scheduling strategy for the current rolling interval.

9. A multi-source microgrid optimization scheduling method according to claim 8, characterized in that: The interval possibility method is used to update the objective function of the day-ahead optimization model. The objective function of the updated day-ahead optimization model is: ， in, is the objective function of the day-ahead optimization model, is the input variable of the day-ahead optimization model, i.e., the parameter data of the current rolling interval. is the control variable of the day-ahead optimization model, i.e., the output power of each distributed power source and energy conversion device in the current rolling interval in the day-ahead optimization scheduling strategy. is the weighted upper limit of each cost of the microgrid, is the weighted lower limit of each cost of the microgrid, is the probability of weighted intervals of each cost of the microgrid; The power balance constraint of the day-ahead optimization model is updated. The updated power balance constraint is: ， in, Indicates the lower limit of load demand, represents the load demand possibility, For wind turbines The corrected output power of the scheduling cycle is: For the The power purchased by the power grid during a dispatch cycle, For the The output power of the battery in a scheduling cycle, Indicates that the area to be optimized is Load demand for a dispatch cycle; The cooling and heating load constraints are: ， in, Indicates that the absorption chiller unit is The cooling capacity of a scheduling cycle, Indicates that the electric refrigeration unit is The cooling capacity of a scheduling cycle, Indicates the possible range of cooling capacity, Indicates The upper limit of cooling capacity for a scheduling cycle, Indicates The lower limit of cooling capacity in a scheduling cycle.

10. A multi-source microgrid optimization dispatching system, characterized in that: include: The model building module obtains the mathematical models of each distributed power source and energy conversion device in the microgrid in the area to be optimized, and constructs the objective function with the weighted minimization of the fuel cost and operation and maintenance cost of the microgrid, the pollutant emission treatment cost of the microgrid and the load control compensation cost of the microgrid as the goal. The power balance constraint, the state of charge constraint of the energy storage, the power generation constraint, the output constraint of each distributed power source and the ramp rate constraint are set as constraints to build the day-ahead optimization model; The model solving module obtains the parameter data of the day-ahead scheduling period and inputs it into the day-ahead optimization model. The day-ahead optimization model is solved by the whale-bat algorithm to obtain the day-ahead optimization scheduling strategy, including: Initialization unit, used to set the number of whales, maximum number of iterations, fusion coefficient, minimum frequency of sound waves emitted by bats, and maximum frequency of sound waves emitted by bats; initialization , randomly generate the initial position and speed of each whale in the search space; calculate the initial fitness of each whale, and record the position and initial fitness of the whale with the lowest fitness; where, Indicates the current iteration number. The initial position of each whale represents a day-ahead optimization scheduling strategy. A random number generation unit, used to generate a random number in the range of [0, 1] for each whale; A judgment unit, used to judge whether the random number generated by the current whale is less than 0.5; A first position updating unit, configured to update the position of the current whale by circling and preying if the random number generated by the current whale is greater than or equal to 0.5; The second position updating unit is used to use the absolute value of the difference between the fitness of the current whale and the lowest fitness in the historical iteration as the judgment coefficient of the current whale if the random number generated by the current whale is less than 0.5; Determine whether the current whale's judgment coefficient is less than 1. If the current whale's judgment coefficient is greater than or equal to 1, use spiral bubble predation to update the current whale's position; if the current whale's judgment coefficient is less than 1, update the current whale's position and speed based on the fusion coefficient, the minimum frequency of the sound waves emitted by the bat, the maximum frequency of the sound waves emitted by the bat, and the current whale's position and speed. The fitness calculation unit is used to calculate the fitness of each whale according to the previous day optimization model and the position of each whale, and record the position and fitness of the whale with the lowest fitness, and update ; Iteration number judgment unit, used to judge Whether the maximum number of iterations is reached, if yes, the position of the whale with the lowest fitness in the historical iteration is output as the day-ahead optimization scheduling strategy, otherwise the operation of the random number generation unit is executed.

Images