Co-generation system scheduling method and device, electronic equipment and medium

Abstract

The present disclosure provides a scheduling method and device of a combined power generation system, an electronic device and a medium, and relates to the technical field of power scheduling, in particular to a power scheduling method based on a deep learning network. The specific implementation scheme is: obtaining predicted output of a new energy unit of the combined power generation system in a first time period and predicted load of the combined power generation system; determining target output corresponding to all units of the combined power generation system when the profit of the combined power generation system is maximum through a target function according to the predicted output and the predicted load; and adjusting actual output of at least one unit in the first time period according to the target output corresponding to the unit. Thus, the supply and demand of the power generation system are balanced, the deviation between the power generation amount and the load of the power generation system is reduced, the economic benefit of the power generation system is improved, the charging and discharging frequency of the energy storage unit is reduced, and the service life of the energy storage unit is improved.

Description

Technical Field

[0001] This disclosure relates to the field of power dispatching technology, specifically to power dispatching methods based on deep learning networks, and particularly to dispatching methods, devices, electronic equipment, and media for combined power generation systems of new energy generating units. Background Technology

[0002] In recent years, to address environmental and energy sustainability issues, countries worldwide have advocated for the use of clean energy, including wind and solar power. Due to the random and intermittent nature of wind and solar energy, independent wind and solar power systems struggle to provide stable and continuous power output. By incorporating wind and solar power complementarity with energy storage and gas-fired power generation, a combined wind-solar-storage-gas power generation system can be formed. This system fully leverages the natural temporal and geographical complementarity of wind and solar energy, while utilizing the charging and discharging capabilities of energy storage and the self-generating function of gas to improve the power output characteristics of the wind-solar complementary system. This mitigates the volatility and intermittency of wind and solar power generation, achieves power supply and demand balance, reduces its adverse impact on the power system, and increases the power system's capacity to absorb and accept renewable energy. Therefore, how to regulate the operation of wind, solar, energy storage, and gas-fired power has become a pressing issue. Summary of the Invention

[0003] This disclosure provides a scheduling method, apparatus, equipment, and storage medium for a combined power generation system.

[0004] According to a first aspect of this disclosure, a method for dispatching a combined power generation system is provided, comprising:

[0005] The predicted output of the new energy units in the combined power generation system and the predicted load of the combined power generation system are obtained in the first time period.

[0006] Based on the predicted output and the predicted load, the target output of all units is determined by an objective function to maximize the profit of the combined power generation system.

[0007] Based on the target output corresponding to the unit, adjust the actual output of at least one of the units during the first time period.

[0008] According to a second aspect of this disclosure, a dispatching device for a combined power generation system is provided, comprising:

[0009] The acquisition module is configured to acquire the predicted output of the new energy units of the combined power generation system and the predicted load of the combined power generation system in the first time period.

[0010] The calculation and analysis module is configured to determine the target output of all units when the profit of the combined power generation system is maximized, based on the predicted output and the predicted load, through an objective function.

[0011] The adjustment module is configured to adjust the actual output of at least one of the generating units during the first time period according to the target output corresponding to the generating unit.

[0012] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method described in the above technical solution.

[0016] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to perform the method described in the above-described technical solution.

[0017] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in any one of the above technical solutions.

[0018] This disclosure obtains the predicted output of new energy generating units and the predicted load of the power generation system, determines the target output of each unit based on the prediction results, reduces the impact of unstable power generation of new energy generating units on the power system, balances the supply and demand of the power generation system, reduces the deviation between power generation and load of the power generation system, and improves the economic efficiency of the power generation system. In addition, it can also reduce the number of charging and discharging cycles of energy storage units and extend their service life.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0020] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0021] Figure 1 This is a diagram of the combined power generation system architecture in the embodiments of this disclosure;

[0022] Figure 2This is a flowchart illustrating the scheduling method steps of the first combined power generation system in this disclosure embodiment;

[0023] Figure 3 This is a flowchart illustrating the scheduling method steps for the second type of combined power generation system in this disclosure.

[0024] Figure 4 This is a schematic diagram of the scheduling method of the combined power generation system in the embodiments of this disclosure;

[0025] Figure 5 This is a schematic diagram of the DQN (Deep Q Networks) network in the embodiments of this disclosure;

[0026] Figure 6 This is a block diagram of the dispatching device of the first type of combined power generation system in this disclosure embodiment;

[0027] Figure 7 This is a block diagram of the dispatching device of the second type of combined power generation system in this disclosure embodiment;

[0028] Figure 8 This is a schematic block diagram of an example electronic device in an embodiment of this disclosure. Detailed Implementation

[0029] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0030] Figure 1 This diagram illustrates a combined wind, solar, energy storage, and gas turbine power generation system architecture. The system comprises wind turbines, photovoltaic generators, energy storage units, and gas turbines, supplying the load required by the power grid using these four energy sources. When calculating the economic benefits of the wind-solar-energy storage power generation system, the power generation of each unit and the power grid load must be considered. The deviation between these two must be minimized, and a certain balance must be maintained between the power generation of each unit and the grid load. If the total power generation of the units exceeds the grid load, the excess energy needs to be stored through energy storage units. However, energy storage units have limited capacity and incur storage costs; furthermore, frequent charging and discharging can shorten the lifespan of the energy storage batteries. Therefore, it is necessary to address the imbalance between wind, solar, and energy storage output and grid load, minimizing the deviation between power generation and grid load to maintain a supply-demand balance.

[0031] To address the aforementioned technical problems, this disclosure provides a scheduling method for a combined power generation system, such as... Figure 2 As shown, it includes:

[0032] Step S201: Obtain the predicted output of the new energy units in the combined power generation system and the predicted load of the combined power generation system in the first time period.

[0033] The combined power generation system disclosed herein includes renewable energy units and conventional units. Renewable energy units may include one or more of the following: wind turbines, photovoltaic power plants, tidal power plants, etc. Due to the random and intermittent nature of renewable energy units, independent wind power systems and / or photovoltaic power systems cannot provide stable and continuous power output. Therefore, conventional units are needed as a supplement when the renewable energy units' power generation is insufficient. Conventional units may include energy storage units, and may also include one or more of the following: gas turbines, coal-fired power plants, etc. A combined power generation system is formed based on the combination of renewable energy units and conventional units. The scheduling method disclosed herein is mainly used to adjust the output of each unit in the short term. The first time period refers to a future time period where the output needs to be predicted and adjusted. For example, if we want to adjust the unit output at 3 PM, we can obtain a prediction result based on historical data before 3 PM, and adjust the actual output of the unit at 3 PM based on this prediction result. In this disclosure, "output" refers to power generation, "predicted output" refers to the predicted power generation of new energy units based on historical data, and "predicted load" refers to the power generation required by the combined power generation system in the first time period.

[0034] Step S202: Based on the predicted output and predicted load, determine the target output of all units when the profit of the combined power generation system is maximized through the objective function.

[0035] Specifically, when controlling the output of each unit in the short term, this disclosure first obtains the predicted output of the new energy units and the predicted load of the distribution network. For example, if the predicted output of the new energy units is 2000 kW and 1000 kW respectively, and the predicted load required by the power grid is 6000 kW, then the power generation of the new energy units cannot meet the predicted load and cannot supply the load of the distribution network. Therefore, the output of the conventional units can be increased. The predicted results are substituted into the objective function, and the target output of each unit when the profit of the power generation system is maximized is calculated through the objective function. Based on the target output, the actual output of each unit is controlled to reduce the impact of unstable power generation of the new energy units, maintain the balance between power grid supply and demand as much as possible, and improve the economic efficiency of the power generation system as much as possible.

[0036] Step S203: Adjust the actual output of at least one unit in the first time period according to the target output of the unit.

[0037] After calculating the target output for each unit using the objective function, the actual output of at least one unit can be adjusted based on the target output. In some cases, not all units need adjustment. For example, if the predicted output of a renewable energy unit is 3000 kW and the predicted load is 6000 kW, the total power generation of the system is insufficient, and only the output of conventional units can be adjusted. Conversely, if the predicted output of a renewable energy unit is 3000 kW and the predicted load is 2000 kW, the power generation system experiences a supply exceeding demand; therefore, the output of the renewable energy unit can be reduced, or the energy storage units can be charged.

[0038] As an optional implementation method, such as Figure 3 As shown, the scheduling method includes:

[0039] Step S204: Update the objective function based on the target output of the unit and the constraints of the unit; wherein the constraints are related to the performance of the unit.

[0040] Constraints related to generator performance mean that the constraints are determined by the generator's physical conditions. For example, if a wind turbine has a rated power of 1000 kW, its maximum output is also 1000 kW. If the calculated target output requires 1500 kW, the generator cannot achieve this, resulting in a power deviation. Therefore, this example pre-sets certain constraints for each generator. If the target output calculated by the objective function does not meet the corresponding constraints, the adjustment command of the objective function will be penalized, promoting continuous updates and optimization of the objective function, improving adjustment accuracy, reducing power deviation, promoting supply and demand balance in the power generation system, and improving economic efficiency.

[0041] As an optional implementation, the new energy units in this embodiment may include wind turbines and photovoltaic units; in addition to new energy units, the units may also include conventional units such as energy storage units and gas turbines. It should be noted that the new energy units are not limited to the wind and solar turbines in this embodiment, but may also include other new energy power generation units such as tidal power generators. Conventional units are also not limited to energy storage units and gas turbines, but may also include coal-fired power generators, etc.

[0042] As an optional implementation, the predicted output includes the predicted wind power output corresponding to the wind turbine and the predicted photovoltaic output corresponding to the photovoltaic unit. Step S201, obtaining the predicted output of the new energy units of the combined power generation system and the predicted load of the combined power generation system in the first time period includes: determining the predicted wind power output in the first time period based on historical wind speed and historical air pressure; determining the predicted photovoltaic output in the first time period based on historical irradiance and historical temperature; and determining the predicted load in the first time period based on historical load.

[0043] like Figure 4 As shown, the historical data in this embodiment refers to data collected before the first time period. Suppose we need to predict the wind turbine output at 3 PM, we can collect historical wind speed and air pressure data from before 3 PM. For example, we can collect wind speed and air pressure data from around 2 PM. Based on this historical data, we can predict the wind turbine output at 3 PM. The predicted output can include the turbine's operating status, i.e., whether the turbine is running. Furthermore, the output of the turbine in its operating state can be represented by a probability distribution, such as [[10,20,30],[0.1,0.2,0.7]]. The former is the predicted probability distribution, and the latter is the probability. For example, at 3 PM, the probability of a wind power output of 10 kW is 0.1, 20 kW is 0.2, and 30 kW is 0.7. We can use the SUM function to sum these probabilities to obtain the predicted wind turbine output. Similarly, when predicting photovoltaic (PV) turbine output, we can also collect irradiance and temperature data from around 2 PM to predict the PV turbine output at 3 PM. When obtaining the predicted load, it is possible to make predictions based on the historical loads of the three points in the past and analyze the load change trends to predict the load at the three points.

[0044] The target output of the generating units includes not only the predicted output of wind power and solar power, but also the predicted output of energy storage units and gas turbine units in the first time period. After obtaining the predicted output of wind power and solar power, the target output of all units is determined through an objective function based on the predicted output and predicted load. This includes determining the predicted output of energy storage and gas turbine units based on the predicted wind power, solar power, and predicted load. In other words, the output of energy storage and gas turbine units is predicted based on the predicted wind and solar power output, thereby reducing the deviation from the predicted load.

[0045] As an optional implementation method, determining the predicted output of energy storage and the predicted output of gas based on the predicted output of wind power, the predicted output of photovoltaic power, and the predicted load includes: determining the sum of the predicted output of energy storage and gas based on the predicted output of wind power, the preset output of photovoltaic power, and the predicted load; obtaining the remaining energy storage capacity and energy storage cost of the energy storage unit and the gas output cost of the gas unit; and allocating the sum of the predicted output of energy storage and gas based on the remaining energy storage capacity, energy storage cost, and gas output cost to obtain the predicted output of energy storage and the predicted output of gas.

[0046] The total predicted output of gas storage refers to the total power generation required by gas storage. It clarifies how much electricity gas storage needs to generate to meet the power system's load, and then allocates the output of energy storage units and gas turbine units based on the cost of gas storage and the current remaining energy storage capacity. The output of energy storage units and gas turbine units is calculated based on the predicted output of wind and solar power. Due to the instability of wind and solar power generation, the output of energy storage units and gas turbine units can serve as a supplement. When wind and solar output exceeds the grid's required load, the excess energy can be stored in energy storage units. When the grid load is high and peak shaving and valley filling are needed, the output of energy storage units can be increased to reduce wind and solar curtailment. Generally, the output cost of wind and solar power is zero. The output of energy storage units is related to the output of gas storage units and can be expressed as a quadratic term relationship: cost = ax 2 The formula is: +bx+c, where a represents the quadratic coefficient, b represents the linear coefficient, c is the constant coefficient, and x represents the power output. Additionally, the current energy storage capacity of the energy storage unit must be considered. When predicting the power output of energy storage units and gas turbine units, not only the difference between the predicted wind and solar power output and the predicted load must be considered, but also the cost of the power output of the energy storage units and gas turbine units. Since energy storage also incurs certain costs, storing more is not necessarily better. When the output of wind and solar power far exceeds the load, it is advisable to consider curtailing a certain amount of wind and solar power.

[0047] In this embodiment, after obtaining the predicted wind power output, predicted photovoltaic output, and predicted load, online rolling optimization is performed on the target based on the prediction results. The optimization of predictive control is not performed offline in a single step, but rather repeatedly online as the sampling time progresses; hence the term "rolling optimization." Although this rolling optimization may not yield an ideal global optimal solution, repeatedly optimizing the deviation at each sampling time allows for timely correction of various complex situations that arise during the control process, ensuring the timeliness of adjustments. Based on the prediction results, the target output within the first time period is continuously corrected, gradually bringing the total output of each unit closer to the actual power generation (actual load) demand of the grid. This reduces grid fluctuations caused by the uncertainty of wind and solar power generation, ensures a more reasonable target output for each unit, maintains the supply-demand balance of the grid, and improves the economic efficiency of the power system.

[0048] As an optional implementation, the objective function includes: determining the total predicted electricity sales price based on the total predicted output of all generating units, the load loss of the combined generation system, and the electricity price; obtaining the output cost corresponding to all generating units; determining a penalty term based on the deviation between the total predicted output of all generating units and the predicted load; and determining the profit of the combined generation system based on the total predicted electricity sales price, the output cost corresponding to all generating units, and the penalty term.

[0049] Specifically, the objective function for maximizing profit can be expressed as: reward = (solar power + wind power + energy storage + gas-fired power – load loss) * electricity price - (solar cost + wind power cost + energy storage cost + gas-fired power cost) - penalty term * (solar power + wind power + energy storage power + gas-fired power – predicted load), where reward represents the profit generated by the power generation system, wind power, solar power, energy storage power, and gas-fired power represent the predicted output of wind, solar, energy storage, and gas-fired power, respectively, and wind power cost, solar power cost, energy storage cost, and gas-fired power cost represent the predicted costs of wind, solar, energy storage, and gas-fired power, respectively. The predicted output of wind and solar power is based on historical data, but as one of the variables, it can also be adjusted through the objective function to ultimately output the target output of each wind, solar, energy storage, and gas-fired power unit. When calculating profit, the predicted total electricity sales price can be subtracted from the costs of wind, solar, gas storage, and fuel oil. Additionally, a penalty term due to the power deviation between the predicted output and predicted load of wind, solar, gas, and gas storage should be calculated to obtain the final profit. The penalty term can be a penalty coefficient between 0 and 1, which can be determined based on the actual application scenario. In this embodiment, a penalty term is set in the objective function `reward`, which can promote continuous updating and optimization of the objective function, thereby improving the accuracy of regulation.

[0050] As an optional implementation method, such as Figure 4 As shown, the constraints include: wind power and photovoltaic constraints, namely the maximum and minimum power output of wind turbines; the ramp-up rate of photovoltaic turbine output; the ramp-up rate of gas turbine output; the maximum and minimum capacity of energy storage units; secondary constraints on energy storage power, namely the ramp-up rate of energy storage unit capacity; and soft constraints on power deviation, namely the deviation range between the total predicted output of wind, solar, gas, and energy storage and the predicted load. If the target output of a unit exceeds the corresponding constraint, it will be penalized. For example, if the objective function yields a target output of 1500 kW for wind turbines, but the maximum power of wind turbines is 1000 kW, then the actual output of wind turbines will not reach 1500 kW, resulting in a total output of 500 kW less than the target output of all units, which will not meet the load demand. Therefore, a corresponding penalty will be imposed on this adjustment to encourage the objective function to be continuously updated and optimized, so that the output target output will get closer and closer to the constraint conditions and gradually no longer exceed the constraint conditions.

[0051] As an optional implementation, step S204, updating the objective function based on the target output of the unit and the constraints corresponding to the unit, includes:

[0052] The target output of each unit is compared with the corresponding constraints to determine whether the target output of each unit meets the corresponding constraints.

[0053] If the target output of any unit does not meet the corresponding constraint, the objective function is subtracted from the first preset value.

[0054] In response to the fact that the target output of all units meets the corresponding constraints and the objective function exceeds the historical maximum value, the objective function is added to the second preset value.

[0055] In response to each iteration of the objective function, the objective function is subtracted from the third preset value.

[0056] For example, the reward and penalty rules in this embodiment may include: when the target output obtained by solving the objective function does not meet the constraint condition (-100), and meets the constraint condition and exceeds the historical maximum reward (+50), the objective function is reduced by 0.1 for each iteration. By establishing a reward and penalty mechanism, positive or negative feedback is given to the action (execution) of the objective function, promoting continuous updating and optimization of the objective function.

[0057] As an optional implementation method, such as Figure 5 As shown, in this embodiment, a DQN network can be used to establish the scheduling objective function. In DQN, we define the loss function as the variance between the target and the predicted values, and we also update the weights to minimize the loss. Initially, the Q-value table and Q-network are randomly initialized, and the subsequent series of predictions are also random. If the action corresponding to the highest Q-value is selected, then this action is naturally also random. At this time, the agent (action initiator) is exploring. As the Q-function converges, the returned Q-values ​​will also tend to be consistent. The Q-learning algorithm is a type of reinforcement learning, which is a method for policy selection. In fact, we can find that the core and training objective of reinforcement learning is to select a suitable policy that maximizes the sum of rewards obtained at the end of each cycle. Reinforcement learning is often used in scenarios that require interaction with the environment. Given a state of the environment, the program selects a corresponding action based on a certain policy. After executing this action, the environment changes, and the state is transformed into a new state S'. After each action is executed, the program receives a reward. The program then adjusts its policy based on the magnitude of the reward to maximize the sum of rewards obtained when all steps are completed and the state reaches the terminal state.

[0058] This disclosure also provides a dispatching device for a wind-solar-storage-gas power generation system, such as Figure 6 As shown, it includes:

[0059] The acquisition module 601 is configured to acquire the predicted output of the new energy units of the combined power generation system and the predicted load of the combined power generation system in the first time period.

[0060] The combined power generation system disclosed herein includes renewable energy units and conventional units. Renewable energy units may include one or more of the following: wind turbines, photovoltaic power plants, tidal power plants, etc. Due to the random and intermittent nature of renewable energy units, independent wind power systems and / or photovoltaic power systems cannot provide stable and continuous power output. Therefore, conventional units are needed as a supplement when the renewable energy units' power generation is insufficient. Conventional units may include energy storage units, and may also include one or more of the following: gas turbines, coal-fired power plants, etc. A combined power generation system is formed based on the combination of renewable energy units and conventional units. The scheduling method disclosed herein is mainly used to adjust the output of each unit in the short term. The first time period refers to a future time period where the output needs to be predicted and adjusted. For example, if we want to adjust the unit output at 3 PM, we can obtain a prediction result based on historical data before 3 PM, and adjust the actual output of the unit at 3 PM based on this prediction result. In this disclosure, "output" refers to power generation, "predicted output" refers to the predicted power generation of new energy units based on historical data, and "predicted load" refers to the power generation required by the combined power generation system in the first time period.

[0061] The calculation and analysis module 602 is configured to determine the target output of all units when the profit of the combined power generation system is maximized by using an objective function based on the predicted output and predicted load.

[0062] Specifically, when controlling the output of each unit in the short term, this disclosure first obtains the predicted output of the new energy units and the predicted load of the distribution network. For example, if the predicted output of the new energy units is 2000 kW and 1000 kW respectively, and the predicted load required by the power grid is 6000 kW, then the power generation of the new energy units cannot meet the predicted load and cannot supply the load of the distribution network. Therefore, the output of the conventional units can be increased. The predicted results are substituted into the objective function, and the target output of each unit when the profit of the power generation system is maximized is calculated through the objective function. The output of each unit is controlled to maintain the balance between power grid supply and demand as much as possible, while maximizing the economic benefits of the power generation system.

[0063] The adjustment module 603 is configured to adjust the actual output of at least one unit in the first time period according to the target output of the unit.

[0064] After the calculation and analysis module 602 calculates the target output of each unit through the objective function, the adjustment module 603 can adjust the actual output of at least one unit according to the target output. In some cases, not all units need to be adjusted. For example, if the predicted output of the renewable energy unit is 3000 kW and the predicted load is 6000 kW, the total power generation of the power generation system is insufficient, and only the output of the conventional units can be adjusted. For example, if the predicted output of the renewable energy unit is 3000 kW and the predicted load is 2000 kW, the power generation system's supply exceeds demand, so the output of the renewable energy unit can be reduced or the energy storage unit can be charged.

[0065] As an optional implementation method, such as Figure 7 As shown, the scheduling device includes:

[0066] The update module 604 is configured to update the objective function based on the target output of the unit and the constraints of the unit; wherein the constraints are related to the performance of the unit.

[0067] Constraints related to generator performance mean that the constraints are determined by the generator's physical conditions. For example, if a wind turbine has a rated power of 1000 kW, its maximum output is also 1000 kW. If the calculated target output requires 1500 kW, the generator cannot achieve this, resulting in a power deviation. Therefore, this example pre-sets certain constraints for each generator. If the target output calculated by the objective function does not meet the corresponding constraints, the adjustment command of the objective function will be penalized, promoting continuous updates and optimization of the objective function, improving adjustment accuracy, reducing power deviation, promoting supply and demand balance in the power generation system, and improving economic efficiency.

[0068] As an optional implementation, the new energy units in this embodiment may include wind turbines and photovoltaic units; in addition to new energy units, the units may also include conventional units such as energy storage units and gas turbines. It should be noted that the new energy units are not limited to the wind and solar turbines in this embodiment, but may also include other new energy power generation units such as tidal power generators. Conventional units are also not limited to energy storage units and gas turbines, but may also include coal-fired power generators, etc.

[0069] As an optional implementation, the predicted output includes the predicted wind power output corresponding to the wind turbine and the predicted photovoltaic output corresponding to the photovoltaic unit. The acquisition module 601 acquires the predicted output of the new energy units of the combined power generation system in the first time period and the predicted load of the combined power generation system, including: determining the predicted wind power output in the first time period based on historical wind speed and historical air pressure; determining the predicted photovoltaic output in the first time period based on historical irradiance and historical temperature; and determining the predicted load in the first time period based on historical load.

[0070] like Figure 4 As shown, the historical data in this embodiment refers to data collected before the first time period. Suppose we need to predict the wind turbine output at 3 PM, we can collect historical wind speed and air pressure data from before 3 PM. For example, we can collect wind speed and air pressure data from around 2 PM. Based on this historical data, we can predict the wind turbine output at 3 PM. The predicted output can include the turbine's operating status, i.e., whether the turbine is running. Furthermore, the output of the turbine in its operating state can be represented by a probability distribution, such as [[10,20,30],[0.1,0.2,0.7]]. The former is the predicted probability distribution, and the latter is the probability. For example, at 3 PM, the probability of a wind power output of 10 kW is 0.1, 20 kW is 0.2, and 30 kW is 0.7. We can use the SUM function to sum these probabilities to obtain the predicted wind turbine output. Similarly, when predicting photovoltaic (PV) turbine output, we can also collect irradiance and temperature data from around 2 PM to predict the PV turbine output at 3 PM. When obtaining the predicted load, it is possible to make predictions based on the historical loads of the three points in the past and analyze the load change trends to predict the load at the three points.

[0071] The target output of the generating units includes not only the predicted output of wind power and solar power, but also the predicted output of energy storage units and gas turbine units in the first time period. After obtaining the predicted output of wind power and solar power, the target output of all units is determined through an objective function based on the predicted output and predicted load. This includes determining the predicted output of energy storage and gas turbine units based on the predicted wind power, solar power, and predicted load. In other words, the output of energy storage and gas turbine units is predicted based on the predicted wind and solar power output, thereby reducing the deviation from the predicted load.

[0072] As an optional implementation method, determining the predicted output of energy storage and the predicted output of gas based on the predicted output of wind power, the predicted output of photovoltaic power, and the predicted load includes: determining the sum of the predicted output of energy storage and gas based on the predicted output of wind power, the preset output of photovoltaic power, and the predicted load; obtaining the remaining energy storage capacity and energy storage cost of the energy storage unit and the gas output cost of the gas unit; and allocating the sum of the predicted output of energy storage and gas based on the remaining energy storage capacity, energy storage cost, and gas output cost to obtain the predicted output of energy storage and the predicted output of gas.

[0073] The total predicted output of gas storage refers to the total power generation required by gas storage. It clarifies how much electricity gas storage needs to generate to meet the power system's load, and then allocates the output of energy storage units and gas turbine units based on the cost of gas storage and the current remaining energy storage capacity. The output of energy storage units and gas turbine units is calculated based on the predicted output of wind and solar power. Due to the instability of wind and solar power generation, the output of energy storage units and gas turbine units can serve as a supplement. When wind and solar output exceeds the grid's required load, the excess energy can be stored in energy storage units. When the grid load is high and peak shaving and valley filling are needed, the output of energy storage units can be increased to reduce wind and solar curtailment. Generally, the output cost of wind and solar power is zero. The output of energy storage units is related to the output of gas storage units and can be expressed as a quadratic term relationship: cost = ax 2 The formula is: +bx+c, where a represents the quadratic coefficient, b represents the linear coefficient, c is the constant coefficient, and x represents the output. Additionally, the current storage capacity of the energy storage unit must be considered. When predicting the output of energy storage and gas turbine units, not only the difference between the predicted wind and solar output and the predicted load must be considered, but also the cost of the energy storage and gas turbine units' output. Since energy storage also incurs costs, storing more is not always better. When the output of wind and solar power far exceeds the load, some wind and solar power can be considered for foreclosure.

[0074] In this embodiment, after obtaining the predicted wind power output, predicted photovoltaic output, and predicted load, online rolling optimization is performed on the target based on the prediction results. The optimization of predictive control is not performed offline in a single step, but rather repeatedly online as the sampling time progresses; hence the term "rolling optimization." Although this rolling optimization may not yield an ideal global optimal solution, repeatedly optimizing the deviation at each sampling time allows for timely correction of various complex situations that arise during the control process, ensuring the timeliness of adjustments. Based on the prediction results, the target output within the first time period is continuously corrected, gradually bringing the total output of each unit closer to the actual power generation (actual load) demand of the grid. This reduces grid fluctuations caused by the uncertainty of wind and solar power generation, ensures a more reasonable target output for each unit, maintains the supply-demand balance of the grid, and improves the economic efficiency of the power system.

[0075] As an optional implementation, the objective function includes: determining the total predicted electricity sales price based on the total predicted output of all generating units, the load loss of the combined generation system, and the electricity price; obtaining the output cost corresponding to all generating units; determining a penalty term based on the deviation between the total predicted output of all generating units and the predicted load; and determining the profit of the combined generation system based on the total predicted electricity sales price, the output cost corresponding to all generating units, and the penalty term.

[0076] Specifically, the objective function for maximizing profit can be expressed as: reward = (solar power + wind power + energy storage + gas-fired power – load loss) * electricity price - (solar cost + wind power cost + energy storage cost + gas-fired power cost) - penalty term * (solar power + wind power + energy storage power + gas-fired power – predicted load), where reward represents the profit generated by the power generation system, wind power, solar power, energy storage power, and gas-fired power represent the predicted output of wind, solar, energy storage, and gas-fired power, respectively, and wind power cost, solar power cost, energy storage cost, and gas-fired power cost represent the predicted costs of wind, solar, energy storage, and gas-fired power, respectively. The predicted output of wind and solar power is based on historical data, but as one of the variables, it can also be adjusted through the objective function to ultimately output the target output of each wind, solar, energy storage, and gas-fired power unit. When calculating profit, the predicted total electricity sales price can be subtracted from the costs of wind, solar, gas storage, and fuel oil. Additionally, a penalty term due to the power deviation between the predicted output and predicted load of wind, solar, gas, and gas storage should be calculated to obtain the final profit. The penalty term can be a penalty coefficient between 0 and 1, which can be determined based on the actual application scenario. In this embodiment, a penalty term is set in the objective function `reward`, which can promote continuous updating and optimization of the objective function, thereby improving the accuracy of regulation.

[0077] As an optional implementation method, such as Figure 4 As shown, the constraints include: wind power and photovoltaic constraints, namely the maximum and minimum power output of wind turbines; the ramp-up rate of photovoltaic turbine output; the ramp-up rate of gas turbine output; the maximum and minimum capacity of energy storage units; secondary constraints on energy storage power, namely the ramp-up rate of energy storage unit capacity; and soft constraints on power deviation, namely the deviation range between the total predicted output of wind, solar, gas, and energy storage and the predicted load. If the target output of a unit exceeds the corresponding constraint, it will be penalized. For example, if the objective function yields a target output of 1500 kW for wind turbines, but the maximum power of wind turbines is 1000 kW, then the actual output of wind turbines will not reach 1500 kW, resulting in a total output of 500 kW less than the target output of all units, which will not meet the load demand. Therefore, a corresponding penalty will be imposed on this adjustment to encourage the objective function to be continuously updated and optimized, so that the output target output will get closer and closer to the constraint conditions and gradually no longer exceed the constraint conditions.

[0078] As an optional implementation, the update module 604 updates the objective function according to the target output of the unit and the constraints corresponding to the unit, including:

[0079] The target output of each unit is compared with the corresponding constraints to determine whether the target output of each unit meets the corresponding constraints.

[0080] If the target output of any unit does not meet the corresponding constraint, the objective function is subtracted from the first preset value.

[0081] In response to the fact that the target output of all units meets the corresponding constraints and the objective function exceeds the historical maximum value, the objective function is added to the second preset value.

[0082] In response to each iteration of the objective function, the objective function is subtracted from the third preset value.

[0083] For example, the reward and penalty rules in this embodiment may include: when the target output obtained by solving the objective function does not meet the constraint condition (-100), and meets the constraint condition and exceeds the historical maximum reward (+50), the objective function is reduced by 0.1 for each iteration. By establishing a reward and penalty mechanism, positive or negative feedback is given to the action (execution) of the objective function, promoting continuous updating and optimization of the objective function.

[0084] As an optional implementation method, such as Figure 5 As shown, in this embodiment, a DQN network can be used to establish the scheduling objective function. In DQN, we define the loss function as the variance between the target and the predicted values, and we also update the weights to minimize the loss. Initially, the Q-value table and Q-network are randomly initialized, and the subsequent series of predictions are also random. If the action corresponding to the highest Q-value is selected, then this action is naturally also random. At this time, the agent (action initiator) is exploring. As the Q-function converges, the returned Q-values ​​will also tend to be consistent. The Q-learning algorithm is a type of reinforcement learning, which is a method for policy selection. In fact, we can find that the core and training objective of reinforcement learning is to select a suitable policy that maximizes the sum of rewards obtained at the end of each cycle. Reinforcement learning is often used in scenarios that require interaction with the environment. Given a state of the environment, the program selects a corresponding action based on a certain policy. After executing this action, the environment changes, and the state is transformed into a new state S'. After each action is executed, the program receives a reward. The program then adjusts its policy based on the magnitude of the reward to maximize the sum of rewards obtained when all steps are completed and the state reaches the terminal state.

[0085] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0086] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0087] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0088] like Figure 8 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0089] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0090] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning objective function algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the scheduling method for a wind-solar-storage-gas power generation system. For example, in some embodiments, the scheduling method for a wind-solar-storage-gas power generation system can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the scheduling method for a wind-solar-storage-gas power generation system described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform a scheduling method for a wind-solar-storage-gas power generation system by any other suitable means (e.g., by means of firmware).

[0091] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0092] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0093] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0094] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0095] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0096] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0097] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0098] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A dispatching method for a combined power generation system, comprising: The predicted output of the new energy units of the combined power generation system and the predicted load of the combined power generation system are obtained in the first time period. The new energy units include wind turbines and photovoltaic units. Based on the predicted output and the predicted load, the target output of all units is determined by an objective function to maximize the profit of the combined power generation system. Based on the target output of the unit, adjust the actual output of at least one of the units in the first time period; The predicted output includes the predicted wind power output corresponding to the wind turbine and the predicted photovoltaic output corresponding to the photovoltaic unit. Obtaining the predicted output of the new energy units in the combined power generation system during the first time period and the predicted load of the combined power generation system includes: The predicted wind power output for the first time period is determined based on historical wind speed and historical air pressure. The predicted wind power output is obtained by summing the probability distribution of the output of the wind turbine under operating conditions. The photovoltaic predicted output for the first time period is determined based on historical irradiance and historical temperature. The photovoltaic predicted output is obtained by summing the probability distribution of the output of the photovoltaic unit under operating conditions. The predicted load for the first time period is determined based on historical load data. The target output corresponding to the generating units includes the predicted energy storage output of the energy storage units in the first time period and the predicted gas output of the gas turbine units in the first time period; the step of determining the target output corresponding to all generating units when the profit of the combined power generation system is maximized by using an objective function based on the predicted output and the predicted load includes: The total predicted output of fuel storage is determined based on the predicted wind power output, the predicted photovoltaic output, and the predicted load. Obtain the remaining energy storage capacity of the energy storage unit, the energy storage cost, and the gas output cost of the gas turbine unit; The total predicted output of the energy storage and gas is allocated based on the remaining energy storage capacity, the energy storage cost, and the gas output cost to obtain the predicted energy storage output and the predicted gas output; the target output is obtained through rolling optimization. The method further includes: The target output of each unit is compared with the corresponding constraint conditions to determine whether the target output of each unit meets the corresponding constraint conditions. In response to the fact that the target output of any of the units does not meet the corresponding constraint, the objective function is subtracted by a first preset value; In response to the fact that the target output of all the units meets the corresponding constraint conditions and the objective function exceeds the historical maximum value, the objective function is added to the second preset value; In response to each iteration of the objective function, the objective function is subtracted by a third preset value; The objective function is constructed using a deep Q-network and trained and updated through reinforcement learning to maximize the cumulative reward.

2. The method according to claim 1, further comprising: The objective function is updated based on the target output of the unit and the constraints corresponding to the unit; wherein the constraints are related to the performance of the unit.

3. The method of claim 1 or 2, wherein, The new energy units include wind turbines and photovoltaic units; in addition to the new energy units, the units also include energy storage units and gas turbines.

4. The method of any of claims 1-3, wherein, The objective function includes: The total predicted electricity price is determined based on the total predicted output of all the aforementioned units, the load loss of the combined power generation system, and the electricity price. Obtain the output cost corresponding to all the aforementioned units; A penalty term is determined based on the deviation between the total predicted output of all the aforementioned units and the predicted load; The profit of the combined power generation system is determined based on the predicted total electricity sales price, the output cost of all the generating units, and the penalty term.

5. The method according to claim 2, wherein, The constraints include: The maximum and minimum power output of the wind turbine generator; The ramp-up rate of the photovoltaic unit's output; The ramp rate of the gas turbine unit's output; The maximum and minimum capacity of the energy storage unit; The ramp-up rate of the energy storage unit's energy storage capacity; The range of deviation between the total predicted output of all the units and the predicted load.

6. A dispatching device for a combined power generation system, comprising: The acquisition module is configured to acquire the predicted output of the new energy units of the combined power generation system and the predicted load of the combined power generation system during a first time period; the new energy units include wind turbines and photovoltaic units. The calculation and analysis module is configured to determine the target output of all units when the profit of the combined power generation system is maximized, based on the predicted output and the predicted load, through an objective function. The adjustment module is configured to adjust the actual output of at least one of the generating units during the first time period according to the target output corresponding to the generating unit. The predicted output includes the predicted wind power output corresponding to the wind turbine and the predicted photovoltaic output corresponding to the photovoltaic unit. The acquisition module is specifically used for: The predicted wind power output for the first time period is determined based on historical wind speed and historical air pressure. The predicted wind power output is obtained by summing the probability distribution of the output of the wind turbine under operating conditions. The photovoltaic predicted output for the first time period is determined based on historical irradiance and historical temperature. The photovoltaic predicted output is obtained by summing the probability distribution of the output of the photovoltaic unit under operating conditions. The predicted load for the first time period is determined based on historical load data. The target output of the unit includes the predicted energy storage output of the energy storage unit in the first time period and the predicted gas output of the gas turbine unit in the first time period. The calculation and analysis module determines the target output of all units when the profit of the combined power generation system is maximized, based on the predicted output and the predicted load, through an objective function, including: The total predicted output of gas storage is determined based on the predicted output of wind power, the predicted output of photovoltaic power, and the predicted load; the remaining energy storage capacity and energy storage cost of the energy storage unit and the gas output cost of the gas turbine unit are obtained; the total predicted output of gas storage is allocated based on the remaining energy storage capacity, the energy storage cost, and the gas output cost to obtain the predicted output of energy storage and the predicted output of gas, wherein the target output is obtained through rolling optimization; The update module is configured as follows: The target output of each unit is compared with the corresponding constraint conditions to determine whether the target output of each unit meets the corresponding constraint conditions. In response to the fact that the target output of any of the units does not meet the corresponding constraint, the objective function is subtracted by a first preset value; In response to the fact that the target output of all the units meets the corresponding constraint conditions and the objective function exceeds the historical maximum value, the objective function is added to the second preset value; In response to each iteration of the objective function, the objective function is subtracted by a third preset value; The objective function is constructed using a deep Q-network and trained and updated through reinforcement learning to maximize the cumulative reward.

7. The scheduling device according to claim 6, further comprising: The update module is configured to update the objective function based on the target output of the unit and the constraints of the unit; wherein the constraints are related to the performance of the unit.

8. The dispatching device according to claim 6 or 7, wherein the new energy unit includes a wind turbine and a photovoltaic unit; the unit further includes an energy storage unit and a gas turbine in addition to the new energy unit.

9. The scheduling device according to any one of claims 6-8, wherein, The objective function includes: The total predicted electricity price is determined based on the total predicted output of all the aforementioned units, the load loss of the combined power generation system, and the electricity price. Obtain the output cost corresponding to all the aforementioned units; A penalty term is determined based on the deviation between the total predicted output of all the aforementioned units and the predicted load; The profit of the combined power generation system is determined based on the predicted total electricity sales price, the output cost of all the generating units, and the penalty term.

10. The scheduling device according to claim 7, wherein, The constraints include: The maximum and minimum power output of the wind turbine generator; The ramp-up rate of the photovoltaic unit's output; The ramp rate of the gas turbine unit's output; The maximum and minimum capacity of the energy storage unit; The ramp-up rate of the energy storage unit's energy storage capacity; The range of deviation between the total predicted output of all the units and the predicted load.

11. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

12. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.

13. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-5.

Images