Economic low-carbon optimized operation method and system for AI and mechanism fused energy system

Abstract

The invention relates to the technical field of energy-carbon optimization of a source-network-load-storage integrated energy system, and provides an economic low-carbon optimization operation method and system for an AI and mechanism fused energy system, and the method comprises the steps: constructing a detailed mathematical model of source-side, network-side, load-side and storage-side AI and mechanism fusion containing multiple types of objects; then, establishing a source-grid-load-storage integrated energy system model in which an alternating-current power grid and a direct-current power grid are coupled and interconnected; secondly, establishing a day-ahead economic low-carbon optimization operation model; and finally, by setting intra-day and day-ahead coupling nested boundary conditions, establishing an intra-day economic low-carbon optimization operation model, performing multi-objective optimization solution, and outputting integrated energy system information. The AI and mechanism fused AC / DC power grid coupling interconnection source network load storage integrated energy system detailed model is constructed, economic and low-carbon operation of the integrated energy system is facilitated, coordinated and stable operation of the source network load storage is facilitated, popularization and application of AC / DC power grid coupling interconnection are accelerated, and deep fusion of the AI model and the mechanism model is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of energy-carbon optimization of an integrated source-grid-load-storage energy system integrating AI with a mechanism model, and in particular to a method and system for economical and low-carbon optimized operation of an energy system integrating AI with a mechanism model. Background Art

[0002] As the global energy structure transition progresses, integrated energy systems with power generation, grid loads, and storage, as a crucial component of the new power system, face significant challenges in optimizing both economic efficiency and low carbon performance. Traditional energy system operation methods primarily rely on mechanistic models. While these models can provide theoretical guidance, they struggle to address the significant uncertainty associated with renewable energy integration and often suffer from issues such as insufficient model accuracy and poor adaptability. AI (artificial intelligence) technology possesses powerful data processing and prediction capabilities. By learning and analyzing historical data, it can uncover underlying operational patterns and provide new pathways for optimizing energy system operation. However, relatively little research has yet to fully integrate AI technology with mechanistic models for application in integrated energy systems with power generation, grid loads, and storage. AI methods that rely solely on data-driven methods lack physical interpretability and are unable to meet the fundamental requirements for the safe operation of new energy systems.

[0003] Currently, integrated energy systems involving power generation, grid loading, and storage involve multiple types of equipment and systems, including various types of power generation equipment, AC / DC power grids, multiple load types, and diverse energy storage devices. Coordinated regulation and optimized operation across multiple timescales are crucial. However, current operational methods are inadequate in terms of coordinated regulation and optimized operation between equipment and AC / DC power grids, failing to fully leverage the strengths of each component and achieve overall system optimization. The existing technologies mainly have the following deficiencies: (1) The coupling mechanism of AC / DC power grids is complex, the existing AC / DC power grid coupling model is overly simplified, and the traditional energy balance modeling method is difficult to accurately describe the interactive characteristics of AC / DC power grids. The existing model does not fully consider key factors such as flow mutual assistance and the location of sources, grids, loads and storage at different nodes of AC / DC power grids, which affects the feasibility of the optimization results; (2) The accuracy and adaptability of traditional mechanism models are insufficient in complex and uncertain operating environments. The existing methods fail to effectively integrate the advantages of AI with the physical interpretability of mechanism models, resulting in difficulty in balancing prediction accuracy and operational safety; (3) The day-ahead and intraday multi-time scale optimization lacks an effective coupling mechanism, and its ability to cope with the uncertainty of renewable energy generation is limited. It is difficult to adapt to the real-time fluctuations of renewable energy output and load demand. The existing operating method is difficult to achieve global optimization operation of the system, and a single economic optimization method cannot meet the needs of economic and low-carbon operation of the energy system. Summary of the Invention

[0004] The purpose of the present invention is to solve at least one technical problem in the background technology and provide an energy system economic low-carbon optimization operation method and system that integrates AI and mechanism.

[0005] To achieve the above objectives, the present invention provides an energy system economic low-carbon optimization operation method that integrates AI and mechanism, comprising: Establish source-side models, including large-scale power grid power supply models and photovoltaic power generation AI prediction models; Establish grid-side models, including AC grid power flow model and DC distribution network power balance model; Establish a load-side model, including a water pump motor model and an uncontrollable power load prediction model; Establish a storage-side model, including power storage facility model and water storage facility model; The source side model, grid side model, load side model and storage side model are coupled and interconnected to form an integrated energy system model; Set up a day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; Setting a daily economic low-carbon optimization operation strategy for the integrated energy system model, and setting coupling connection conditions between the daily economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed. The integrated energy system information is input into the integrated energy system, and the integrated energy system economic and low-carbon optimization result information is output through the integrated energy system.

[0006] According to one aspect of the present invention, the expression of the large power grid power supply model is as follows: ; Where: The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The upper limit power of the power supply capacity of the grid-connected point tie line power grid to the source-grid-load-storage integrated energy system; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The power upper limit of the capacity of the power returned to the grid by the source-grid-load-storage integrated energy system of the grid connection line; The photovoltaic power generation AI prediction model is a sequence-to-sequence model architecture based on a long short-term memory network combined with an attention mechanism to process temporal dependency models, including: an encoder part model and a decoder part model; The expression of the encoder part model is as follows: ; Where: is the output of the forget gate at time t; is the input gate output at time t; is the output gate output at time t; is the candidate cell state at time t; is the current cell state at time t; for Current cell state at any moment; is the hidden state at time t; for Always hide the status; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; 、 、 、 are all weight matrices, corresponding to the weights of the forget gate, input gate, output gate and cell state respectively; 、 、 、 are all bias vectors, corresponding to the bias items of the forget gate, input gate, output gate and cell state respectively; For variables Sigmoid function; For variables The hyperbolic tangent function of The expression of the decoder part model is as follows: ; Where: Hidden state of the decoder at time t, storing decoding process information; For decoder Always hide the state and store the decoding process information; For the previous step The predicted output at the moment is the normalized photovoltaic power value; z is the context vector, which is initially the final hidden state of the encoder; is the attention weight vector, used to calculate the attention score; For variables The hyperbolic tangent function of 、 Both are attention weight matrices, which handle the trainable parameters of the decoder and encoder states respectively; is the hidden state of the encoder at step i; is the unnormalized attention score, which is the correlation strength between the decoder state at time t and the encoder state at time i; is the unnormalized attention score, which is the strength of the association between the decoder state at time t and the encoder state at time j; is the attention weight, the importance of the encoder at moment i to the decoder at moment t; is the dynamic context vector at time t; To predict the output at time t, it is necessary to denormalize to obtain the actual power value; is the output layer weight matrix, mapping the concatenated state to the output; is the output layer bias, which predicts the output bias term; T is the encoder sequence length, that is, the total number of input time steps; It is a complete long short-term memory network unit calculation process function, which contains the input gate that controls the inflow of new information, the forget gate that controls the retention of historical information, the output gate that controls the state output, and the cell state of the long-term memory carrier; The expression of the photovoltaic power generation mechanical model is as follows: ; Where: is the photovoltaic power generation power at time t; The comprehensive conversion efficiency of the photovoltaic power generation system; The total effective surface area of ​​the photovoltaic power generation system battery components; is the solar radiation intensity at time t; The conversion coefficient between photovoltaic power generation and temperature adjustment; is the working environment temperature of the photovoltaic solar panel at time t; The operating temperature of solar panels under standard test conditions for photovoltaic power generation; is the photovoltaic panel tilt angle correction function; is the tilt angle of the photovoltaic panel; 、 for Different parameters of the function; is the weather type influence coefficient, where Indicates the sunny day coefficient, which is usually set to 1. Indicates the cloudy coefficient, which is generally set to a value between 0.6 and 0.8. Indicates the rainy day coefficient, which is generally set to a value between 0.2 and 0.5; is the weather type at time t; Convert weather type codes and convert text values ​​at different times into numerical values; The expression of the input model of the long short-term memory network time series prediction model is as follows: ; Where: is the solar radiation intensity at time t; is the working environment temperature of photovoltaic power generation at time t; is the relative humidity at time t; is the wind speed at time t; is the weather type at time t; for Photovoltaic power generation at any moment; for Photovoltaic power generation at any moment; 、 、 、 They are minute, hour, day, and month time labels respectively; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; The expression of the output prediction target model of the long short-term memory network time series prediction model is as follows: ; Where: It is a mapping function of a complete long short-term memory network neural network to the time series, realizing the input and output mapping function, including the encoder time series and decoder prediction generation process model content; T is the encoder sequence length, that is, the total number of input time steps; n is the total number of historical feature vectors; 、 、 They are time, time, The predicted photovoltaic power generation power at the moment; 、 、 They are time t, time, The input feature vector at the time, including the standardized PV power generation meteorological data and historical power; According to one aspect of the present invention, the expression of the AC power grid flow model is as follows: ; Where: is the total active power injected into the AC grid node j at time t; is the total reactive power injected into the AC grid node j at time t; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Current amplitude; AC grid line resistance; AC grid line reactance; is the conductance of AC grid node j; is the susceptance of AC grid node j; is the voltage amplitude of AC grid node j at time t; is the voltage amplitude of the AC grid node i at time t; is the set of first nodes in the AC power grid with j as the last node, and Indicates that the first node i is connected to the last node j; is the set of tail nodes in the AC power grid with j as the first node, and Indicates that the tail node i is connected to the head node j; The expression of the DC distribution network power balance model is as follows: ; Where: is the net active power injected into the DC distribution network node i at time t; is the active power flowing from node i to node j in the DC distribution network at time t; The power branch formed by the DC distribution network node i and node j at time t Power loss; is the voltage of DC distribution network node i at time t; is the voltage of DC distribution network node j at time t; is the conductance between nodes i and j in the DC distribution network; Node j is directly connected to node i in the DC distribution network.

[0007] According to one aspect of the present invention, the water pump motor model includes: a fully autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, and a set optimization water pump motor operation mode model, and the operation mode of each model is respectively recorded as The operating mode, The operating mode, In actual operation, an operation mode is selected for optimization. The setting of the operation mode meets the following conditions: ; Where: Select a variable for the fully autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it means that the water pump motor model operates using the fully autonomous optimization water pump motor operation mode model. When the value is 0, it means that the water pump motor model does not operate using the fully autonomous optimization water pump motor operation mode model. A variable is selected for the semi-autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it indicates that the water pump motor model operates using the semi-autonomous optimization water pump motor operation mode model; when the value is 0, it indicates that the water pump motor model does not operate using the semi-autonomous optimization water pump motor operation mode model; The variable is selected for the set-type optimized water pump motor operation mode model, and the value is 0 or 1. When the value is 1, it means that the water pump motor model is operated using the set-type optimized water pump motor operation mode model. When the value is 0, it means that the water pump motor model is not operated using the set-type optimized water pump motor operation mode model. The specific expression of the fully autonomous optimization water pump motor operation mode model is as follows: ; Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The specific expression of the semi-autonomous optimization water pump motor operation mode model is as follows: ; Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; The specific expression of the set-type optimization water pump motor operation mode model is as follows: ; Where: is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; The uncontrollable power load prediction model predicts the uncontrollable power load prediction data point results, including: Get 6 uncontrollable power load data points in the recent history, which are recorded as 、 、 、 、 、 ; The next predicted uncontrollable power load data point is recorded as ; Set the moving average calculation step size is 3; Calculate the first moving average. The expression of the first moving average calculation model is as follows: ; Where: is the result of the first moving average calculation at time t; 、 、 They are time t, time, The historical uncontrollable power load data points at the moment; The step size for moving average calculation is set; After calculating the first moving average, the data sequence is formed: 、 、 、 ; Calculate the second moving average, and the expression of the calculation model is as follows: ; Where: for The second moving average calculation result of the time; 、 、 They are time t, time, The first moving average calculation result of the moment; The step size for moving average calculation is set; To predict the next latest data point of uncontrollable power load, the prediction model expression is as follows:

[0008] Where: To predict the next latest data point of uncontrollable power load, that is, the predicted The data point at that moment, in particular, when hour, Indicates the first latest data point for predicting future uncontrollable power load; is the result of the first moving average calculation at time t; is the result of the second moving average calculation at time t; A new sequence of uncontrollable power load data points is formed, which is recorded as 、 、 、 、 、 ; Based on the new sequence of uncontrollable power load data points, predict the future ; According to the need of predicting the future time period, predict the 、 and other data points within the time period, thereby obtaining the results of all data points of uncontrollable power load prediction.

[0009] According to one aspect of the present invention, the expression of the electricity storage facility model is as follows: ; Where: 、 are the stored energy of the storage facility at time t and at time t, respectively; is the loss rate of the charging and discharging process of the electricity storage facility; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 are the charging efficiency and discharging efficiency of the power storage facility respectively; To optimize the running step resolution; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 They are the upper limit of discharge power and charging power of power storage facilities respectively; 、 They are the lower limit of discharge power regulation and protection of the power storage facility and the lower limit of charging power regulation and protection; 、 They are the upper limit coefficient and lower limit coefficient of the real-time storage energy of the power storage facility respectively; The installed rated capacity of the electricity storage facility; 、 Optimize the storage energy at the beginning and end of the operation cycle of the storage facility respectively; In order to optimize the start and end of the operation cycle, the balance coefficient of the energy storage facility at the start and end states is set. The value range is between 0 and 1. When the value is 0, it means that there is no requirement for the balance between the start and end states; when the value is 1, it means that a complete balance is required, and the start and end states are completely consistent. To optimize the operating cycle; The expression of the water storage facility model is as follows: ; Where: The operating power of the water pump motor at time t; is the functional relationship of the pump motor with respect to the water inlet volume flow rate; is the density of water; is the acceleration due to gravity; is the pump motor head at time t, including static head and pipeline loss; is the water inlet volume flow rate of the pump motor at time t; is the comprehensive energy conversion efficiency of the pump motor at time t, including motor and hydraulic efficiency; 、 、 Different parameters for the functional relationship between the pump motor head and the water inlet volume flow rate; 、 、 Different parameters for the functional relationship between the comprehensive efficiency of energy conversion of the water pump motor and the volume flow rate of water inlet; for The water storage capacity of water storage facilities at any given time; is the water storage capacity of the water storage facility at time t; 、 They are the upper and lower limits of water storage capacity of water storage facilities respectively; is the outlet volume flow rate of the water storage facility at time t; The lower limit of the water discharge volume flow rate during the permitted water discharge period of the water storage facility; The maximum upper limit of the water outlet volume flow rate of the water storage facility; is the upper limit of the water inlet volume flow rate for safe operation of the water pump motor at time t; 、 They are respectively the permitted water-discharge periods and non-water-discharge periods set for water storage facilities during operation and maintenance.

[0010] According to one aspect of the present invention, the integrated energy system model includes: a converter model, a DC bus power balance model, an AC grid power node model, an AC grid safe operation constraint model, and a DC grid safe operation constraint model; The expression of the converter model is as follows: ; Where: The efficiency of the converter in converting DC to AC; The efficiency of the converter in converting AC to DC; is the DC side input active power when the converter converts DC to AC working mode at time t; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; is the average power factor angle when the converter is working; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; is the tangent function; The expression of the DC bus power balance model is as follows: ; Where: is the net active power injected into the DC distribution network node i at time t; The predicted photovoltaic power generation power at time t; is the power actually consumed by photovoltaic power generation in optimized operation at time t; 、 are the discharge power and charging power of the power storage facility at time t respectively; is the DC side input active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; The expression of the AC power grid node model is as follows: ; Where: is the total active power injected into the AC grid node j at time t; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the data point at time t for predicting uncontrollable power load; To increase the total operating power of the water system at time t; is the total reactive power injected into the AC grid node j at time t; is the reactive power of the node at the tie line of the grid connection point at time t; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; It is the average power factor angle when the uncontrollable power load is working; is the average power factor angle when the water pump motor is working; is the tangent function; The expression of the AC power grid safe operation constraint model is as follows: ; Where: is the voltage amplitude of the AC grid node i at time t; is the maximum upper limit of the voltage amplitude of the AC grid node i; is the minimum lower limit of the voltage amplitude of the AC grid node i; is the voltage amplitude of AC grid node j at time t; is the maximum upper limit of the voltage amplitude at node j in the AC grid; is the minimum lower limit of the voltage amplitude of the AC grid node j; is the AC power grid line at time t Current amplitude; AC grid line The maximum upper limit of the current amplitude; AC grid line Minimum lower limit of current amplitude; is the total active power injected into the AC grid node j at time t; The maximum upper limit of the total active power injected into the AC grid node j, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the total active power injected into the AC grid node j, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the total reactive power injected into the AC grid node j at time t; is the maximum upper limit of the total reactive power injected into the AC grid node j; is the minimum lower limit of the total reactive power injected into the AC grid node j; The expression of the DC grid safe operation constraint model is as follows: ; Where: is the voltage of DC distribution network node i at time t; is the maximum upper limit of the voltage at the DC grid node i; is the minimum lower limit of the voltage at the DC grid node i; is the voltage of DC distribution network node j at time t; is the maximum upper limit of the voltage at node j in the DC grid; is the minimum lower limit of the voltage at node j in the DC grid; is the net active power injected into the DC distribution network node i at time t; The maximum upper limit of the net active power injected into the DC distribution network node i, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the net active power injected into the DC distribution network node i, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the active power flowing from node i to node j in the DC distribution network at time t; is the maximum upper limit of active power flowing from node i to node j in the DC distribution network; is the minimum lower limit of active power flowing from node i to node j in the DC distribution network.

[0011] According to one aspect of the present invention, the day-ahead economic low-carbon optimized operation strategy includes: an economic optimized operation model, a low-carbon optimized operation model, and an economic low-carbon optimized operation model; The expression of the economic optimization operation model is as follows: ; Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The electricity price purchased from the grid by the integrated energy system of source, grid, load and storage at time t; The electricity price sold by the integrated energy system of source, grid, load and storage to the large power grid at time t; The expression of the low-carbon optimization operation model is as follows: ; Where: To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; is the power actually consumed by photovoltaic power generation in optimized operation at time t; The unit power equivalent carbon emission coefficient of electricity purchased from the large power grid for the integrated energy system of source, grid, load and storage; The expression of the economic low-carbon optimization operation model is as follows: ; Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the economic and low-carbon operation goals of the integrated energy system of source, grid, load and storage during the operation cycle; 、 、 、 They are respectively the positive deviation weight of economic optimization target, the negative deviation weight of economic optimization target, the positive deviation weight of low-carbon optimization target, and the negative deviation weight of low-carbon optimization target; 、 、 、 They are respectively the positive deviation of economic optimization target, the negative deviation of economic optimization target, the positive deviation of low-carbon optimization target, and the negative deviation of low-carbon optimization target; Acceptable tolerance coefficient for economic optimization objectives; Acceptable tolerance coefficient for low carbon optimization goals; It is the optimal result for single-objective economic optimization; This is the optimal result for single-objective low-carbon optimization.

[0012] According to one aspect of the present invention, the integrated energy system model is configured with an intraday economic low-carbon optimization operation strategy, and coupling connection conditions between the intraday economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy are configured. After the coupling connection conditions are configured, an integrated energy system is formed, including: Solve the day-ahead economic low-carbon optimization operation model and obtain the optimal solution. The optimal solution includes: the power value of the grid connection line at time t supplied to the source-grid-load-storage integrated energy system is recorded as The power returned to the large power grid by the source-grid-load-storage integrated energy system at time t is recorded as , the discharge power of the storage facility at time t is recorded as , the charging power of the storage facility at time t is recorded as The total operating power of the water lifting system at time t is recorded as The outflow volume flow rate of the water storage facility at time t is recorded as , when the converter switches from AC to DC working mode at time t, the AC side input active power is recorded as When the converter converts DC to AC working mode at time t, the AC side output active power is recorded as The economic and low-carbon optimization operation target of the source-grid-load-storage integrated energy system under the optimized operation cycle is recorded as ; Set the daily economic low-carbon optimization operation cycle to ; According to the photovoltaic power generation AI prediction model, the daily economic low-carbon optimization operation cycle is re-predicted as The photovoltaic power generation power at time t is recorded as ,use Replace the photovoltaic power generation forecast power at time t in the economic low-carbon optimization operation of the day before ; According to the uncontrollable power load forecasting model, the daily economic low-carbon optimization operation cycle is re-forecasted as The uncontrollable power load power at time t is recorded as ,use Replace the t-time data point of the uncontrollable power load predicted in the day-ahead economic low-carbon optimization operation ; The daily economic low-carbon optimization operation cycle is On this basis, the economic low-carbon optimization operation model is re-solved, and the intraday day-ahead coupled nested boundary conditions are additionally set. The expression of the intraday day-ahead coupled nested boundary condition model is as follows: ; Where: To obtain the power value of the grid connection line supplied to the source-grid-load-storage integrated energy system at time t after solving the economic low-carbon optimization operation model of the day before; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The fluctuation tolerance coefficient of power supply power is nested and optimized for intraday and day-ahead coupling; To obtain the total operating power of the water lifting system at time t after solving the day-ahead economic low-carbon optimization operation model; To increase the total operating power of the water system at time t; In order to improve the total power of the water system, the day-ahead coupled nested optimization operation fluctuation tolerance coefficient is adopted; To solve the day-ahead economic low-carbon optimization operation model, the grid connection line is obtained at time t and the power returned to the large power grid by the source-grid-load-storage integrated energy system; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient of the returned power is used; To obtain the AC side input active power when the converter switches from AC to DC working mode at time t after solving the day-ahead economic low-carbon optimization operation model; is the AC side input active power when the converter switches from AC to DC working mode at time t; The permissible fluctuation coefficient of the intraday and day-ahead coupled nested optimization operation of the AC side input active power; To solve the economic low-carbon optimization operation model obtained after the day before, the AC side output active power of the converter when the DC is converted to AC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; The permissible fluctuation coefficient of the AC side output active power is determined by nested optimization of intraday and day-ahead coupling operation; To obtain the economic low-carbon optimization operation target value of the source-grid-load-storage integrated energy system under the optimization operation cycle after solving the day-ahead economic low-carbon optimization operation model; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient within the low-carbon optimization operation target day; To obtain the discharge power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the discharge power of the storage facility at time t; The discharge power intraday day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the charging power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the charging power of the power storage facility at time t; The charging power intraday and day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the outflow volume flow of the water storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the outlet volume flow rate of the water storage facility at time t; It is the permissible coefficient of fluctuation of the outlet volume flow rate during the day and the day before the day coupled nested optimization operation.

[0013] To achieve the above objectives, the present invention further provides an energy system economic low-carbon optimization operation system integrating AI and mechanism, comprising: Source-side model building module, which builds the source-side model, including the large-scale power grid power supply model and the photovoltaic power generation AI prediction model; Grid-side model building module, which builds the grid-side model, including the AC grid power flow model and the DC distribution network power balance model; Load-side model building module, which builds the load-side model, including the water pump motor model and the uncontrollable power load prediction model; The reservoir side model establishment module establishes the reservoir side model, including the power storage facility model and the water storage facility model; An integrated energy system model building module couples and interconnects the source side model, grid side model, load side model, and storage side model to form an integrated energy system model; The day-ahead economic low-carbon optimization operation strategy setting module sets the day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; The intraday economic low-carbon optimization operation strategy setting module sets the intraday economic low-carbon optimization operation strategy for the integrated energy system model, and sets the coupling connection conditions between the intraday economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed; The result output module inputs the integrated energy system information into the integrated energy system, and outputs the economic and low-carbon optimization result information of the integrated energy system through the integrated energy system.

[0014] To achieve the above-mentioned objectives, the present invention also provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the method for economic and low-carbon optimization operation of an energy system integrating AI and mechanism as described above is implemented.

[0015] To achieve the above-mentioned objectives, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the economic and low-carbon optimization operation method of the energy system integrating AI and mechanism as described above is implemented.

[0016] According to the solution of the present invention, the present invention balances economic benefits and emission reduction benefits through the optimization of electricity purchase and sales strategies and carbon emission sensitive equipment, enhances the ability of the integrated energy system of source, grid, load and storage to cope with the uncertainty of new energy power generation, and fully taps the potential of renewable energy; the present invention improves the operating stability and safety of source, grid, load and storage equipment and AC and DC power grids through comprehensive modeling of source, grid, load and storage and AC and DC power grid safety constraints, optimizes the coordinated regulation and operation strategies between equipment and AC and DC power grids, improves the overall performance of the system, realizes the global optimized operation of the energy system, and provides strong support for the economic and low-carbon operation of the system; the present invention can accurately reflect the actual operation of the energy system through the deep fusion model of AI and mechanism constructed, and provide a more reliable basis for optimized operation. It not only realizes theoretical and method innovation, but also provides a practical technical solution for the coordinated operation of the integrated energy system of source, grid, load and storage under the background of new power system through the engineerable model architecture and solution algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart schematically illustrates an energy system economic low-carbon optimization operation method integrating AI and mechanism according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The present invention will now be discussed with reference to exemplary embodiments. It should be understood that the embodiments discussed are only intended to enable those skilled in the art to better understand and implement the present invention, rather than to imply any limitation on the scope of the present invention.

[0019] As used herein, the term "including" and variations thereof are to be interpreted as open-ended terms meaning "including, but not limited to." The term "based on" is to be interpreted as "based, at least in part, on." The terms "one embodiment" and "an embodiment" are to be interpreted as "at least one embodiment."

[0020] Figure 1 The flowchart schematically shows the method for economic low-carbon optimization operation of energy system by integrating AI and mechanism according to one embodiment of the present invention. Figure 1 As shown, the energy system economic low-carbon optimization operation method integrating AI and mechanism is characterized by including: Establish source-side models, including large-scale power grid power supply models and photovoltaic power generation AI (artificial intelligence) prediction models; Establish grid-side models, including AC grid power flow model and DC distribution network power balance model; Establish a load-side model, including a water pump motor model and an uncontrollable power load prediction model; Establish a storage-side model, including power storage facility model and water storage facility model; The source side model, grid side model, load side model and storage side model are coupled and interconnected to form an integrated energy system model; Set up a day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; Setting a daily economic low-carbon optimization operation strategy for the integrated energy system model, and setting coupling connection conditions between the daily economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed. The integrated energy system information is input into the integrated energy system, and the integrated energy system economic and low-carbon optimization result information is output through the integrated energy system.

[0021] Furthermore, according to one embodiment of the present invention, the expression of the large power grid power supply model is as follows: ; Where: The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The upper limit power of the power supply capacity of the grid-connected point tie line power grid to the source-grid-load-storage integrated energy system; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The grid connection line is the upper limit power of the power capacity returned to the grid by the source-grid-load-storage integrated energy system; in the large grid power supply model, when the source-grid-load-storage integrated energy system is interconnected with the large grid, there are two grid connection modes, one is to purchase electricity from the large grid and return power to the large grid; the other is to only purchase electricity from the large grid, with anti-backflow equipment installed, and cannot return power to the large grid. At this time, the large grid power supply model Set the value to 0; The photovoltaic power generation AI prediction model is a sequence-to-sequence model architecture based on a long short-term memory network combined with an attention mechanism to process temporal dependency models. It includes an encoder model and a decoder model. The expression of the encoder part model is as follows: ; Where: is the output of the forget gate at time t; is the input gate output at time t; is the output gate output at time t; is the candidate cell state at time t; is the current cell state at time t; for Current cell state at any moment; is the hidden state at time t; for Always hide the status; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; 、 、 、 are all weight matrices, corresponding to the weights of the forget gate, input gate, output gate and cell state respectively; 、 、 、 are bias vectors, corresponding to the bias terms of the forget gate, input gate, output gate and cell state respectively; Sigmoid function; For variables The hyperbolic tangent function of The expression of the decoder part model is as follows: ; Where: Hidden state of the decoder at time t, storing decoding process information; For decoder Always hide the state and store the decoding process information; For the previous step The predicted output at the moment is the normalized photovoltaic power value; z is the context vector, which is initially the final hidden state of the encoder; is the attention weight vector, used to calculate the attention score; For variables The hyperbolic tangent function of 、 Both are attention weight matrices, which handle the trainable parameters of the decoder and encoder states respectively; is the hidden state of the encoder at step i; is the unnormalized attention score, which is the correlation strength between the decoder state at time t and the encoder state at time i; is the unnormalized attention score, which is the strength of the association between the decoder state at time t and the encoder state at time j; is the attention weight, the importance of the encoder at moment i to the decoder at moment t; is the dynamic context vector at time t; To predict the output at time t, it is necessary to denormalize to obtain the actual power value; is the output layer weight matrix, mapping the concatenated state to the output; is the output layer bias, which predicts the output bias term; T is the encoder sequence length, that is, the total number of input time steps; It is a complete long short-term memory network unit calculation process function, which contains the input gate that controls the inflow of new information, the forget gate that controls the retention of historical information, the output gate that controls the state output, and the cell state of the long-term memory carrier; The expression of the photovoltaic power generation mechanism model is as follows: ; Where: is the photovoltaic power generation power at time t; The comprehensive conversion efficiency of the photovoltaic power generation system; The total effective surface area of ​​the photovoltaic power generation system battery components; is the solar radiation intensity at time t; The conversion coefficient between photovoltaic power generation and temperature adjustment; is the working environment temperature of the photovoltaic solar panel at time t; The operating temperature of solar panels under standard test conditions for photovoltaic power generation; is the photovoltaic panel tilt angle correction function; is the tilt angle of the photovoltaic panel; 、 for Different parameters of the function; is the weather type influence coefficient, where Indicates the sunny day coefficient, which is usually set to 1. Indicates the cloudy coefficient, which is generally set to a value between 0.6 and 0.8. Indicates the rainy day coefficient, which is generally set to a value between 0.2 and 0.5; is the weather type at time t; Convert weather type codes and convert text values ​​at different times into numerical values; The expression of the input model of the long short-term memory network time series prediction model is as follows: ; Where: is the solar radiation intensity at time t; is the working environment temperature of photovoltaic power generation at time t; is the relative humidity at time t; is the wind speed at time t; is the weather type at time t; for Photovoltaic power generation at any moment; for Photovoltaic power generation at any moment; 、 、 、 They are minute, hour, day, and month time labels respectively; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; The expression of the output prediction target model of the long short-term memory network time series prediction model is as follows: ; Where: It is a mapping function of a complete long short-term memory network neural network to the time series, realizing the input and output mapping function, including the encoder time series and decoder prediction generation process model content; T is the encoder sequence length, that is, the total number of input time steps; n is the total number of historical feature vectors; 、 、 They are time, time, The predicted photovoltaic power generation power at the moment; 、 、 They are time t, time, The input feature vector at the time, including the standardized PV power generation meteorological data and historical power; Furthermore, according to one embodiment of the present invention, the AC power flow model is expressed as follows: ; Where: is the total active power injected into the AC grid node j at time t; is the total reactive power injected into the AC grid node j at time t; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Current amplitude; AC grid line resistance; AC grid line reactance; is the conductance of AC grid node j; is the susceptance of AC grid node j; is the voltage amplitude of AC grid node j at time t; is the voltage amplitude of the AC grid node i at time t; is the set of first nodes in the AC power grid with j as the last node, and Indicates that the first node i is connected to the last node j; is the set of tail nodes in the AC power grid with j as the first node, and Indicates that the tail node i is connected to the head node j; The expression of the DC distribution network power balance model is as follows: ; Where: is the net active power injected into the DC distribution network node i at time t; is the active power flowing from node i to node j in the DC distribution network at time t; The power branch formed by the DC distribution network node i and node j at time t Power loss; is the voltage of DC distribution network node i at time t; is the voltage of DC distribution network node j at time t; is the conductance between nodes i and j in the DC distribution network; Node j is directly connected to node i in the DC distribution network.

[0022] Furthermore, according to one embodiment of the present invention, the water pump motor model includes: a fully autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, and a set optimization water pump motor operation mode model, and the operation mode of each model is recorded as The operating mode, The operating mode, In actual operation, select an operation mode for optimization. The setting of the operation mode meets the following conditions: ; Where: Select a variable for the fully autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it means that the water pump motor model operates using the fully autonomous optimization water pump motor operation mode model. When the value is 0, it means that the water pump motor model does not operate using the fully autonomous optimization water pump motor operation mode model. A variable is selected for the semi-autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it indicates that the water pump motor model operates using the semi-autonomous optimization water pump motor operation mode model; when the value is 0, it indicates that the water pump motor model does not operate using the semi-autonomous optimization water pump motor operation mode model; The variable is selected for the set-type optimized water pump motor operation mode model, and the value is 0 or 1. When the value is 1, it means that the water pump motor model is operated using the set-type optimized water pump motor operation mode model. When the value is 0, it means that the water pump motor model is not operated using the set-type optimized water pump motor operation mode model. The specific expression of the fully autonomous optimization water pump motor operation mode model is as follows: ; Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The specific expression of the semi-autonomous optimization water pump motor operation mode model is as follows: ; Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; The specific expression of the set-type optimization water pump motor operation mode model is as follows: ; Where: is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; The uncontrollable power load forecasting model predicts the uncontrollable power load forecasting data point results, including: Get 6 uncontrollable power load data points in the recent history, which are recorded as ; The next predicted uncontrollable power load data point is recorded as; Set the moving average calculation step size is 3; Calculate the first moving average. The expression of the first moving average calculation model is as follows: ; Where: is the result of the first moving average calculation at time t; They are time t, time, The historical uncontrollable power load data points at the moment; The step size for moving average calculation is set; After calculating the first moving average, the data sequence is formed: ; Calculate the second moving average, and the expression of the calculation model is as follows: ; Where: for The second moving average calculation result of the time; They are time t, time, The first moving average calculation result of the moment; The step size for moving average calculation is set; To predict the next latest data point of uncontrollable power load, the prediction model expression is as follows:

[0023] Where: To predict the next latest data point of uncontrollable power load, that is, the predicted The data point at that moment, in particular, when hour, Indicates the first latest data point for predicting future uncontrollable power load; is the result of the first moving average calculation at time t; is the result of the second moving average calculation at time t; A new sequence of uncontrollable power load data points is formed, which is recorded as ; Based on the new sequence of uncontrollable power load data points, predict the future ; According to the need of predicting the future time period, predict the 、 and other data points within the time period, thereby obtaining the results of all data points of uncontrollable power load prediction.

[0024] Furthermore, according to one embodiment of the present invention, the expression of the power storage facility model is as follows: ; Where: 、 are the stored energy of the storage facility at time t and at time t, respectively; is the loss rate of the charging and discharging process of the electricity storage facility; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 are the charging efficiency and discharging efficiency of the power storage facility respectively; To optimize the running step resolution; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 They are the upper limit of discharge power and charging power of power storage facilities respectively; 、 They are the lower limit of discharge power regulation and protection of the power storage facility and the lower limit of charging power regulation and protection; 、 They are the upper limit coefficient and lower limit coefficient of the real-time storage energy of the power storage facility respectively; The installed rated capacity of the electricity storage facility; 、 Optimize the storage energy at the beginning and end of the operation cycle of the storage facility respectively; In order to optimize the start and end of the operation cycle, the balance coefficient of the energy storage facility at the start and end states is set. The value range is between 0 and 1. When the value is 0, it means that there is no requirement for the balance between the start and end states; when the value is 1, it means that a complete balance is required, and the start and end states are completely consistent. To optimize the operating cycle; The expression of the water storage facility model is as follows: ; Where: The operating power of the water pump motor at time t; is the functional relationship of the pump motor with respect to the water inlet volume flow rate; is the density of water; is the acceleration due to gravity; is the pump motor head at time t, including static head and pipeline loss; is the water inlet volume flow rate of the pump motor at time t; is the comprehensive energy conversion efficiency of the pump motor at time t, including motor and hydraulic efficiency; Different parameters for the functional relationship between the pump motor head and the water inlet volume flow rate; Different parameters for the functional relationship between the comprehensive efficiency of energy conversion of the water pump motor and the volume flow rate of water inlet; for The water storage capacity of water storage facilities at any given time; is the water storage capacity of the water storage facility at time t; 、 They are the upper and lower limits of water storage capacity of water storage facilities respectively; is the outlet volume flow rate of the water storage facility at time t; The lower limit of the water discharge volume flow rate during the permitted water discharge period of the water storage facility; The maximum upper limit of the water outlet volume flow rate of the water storage facility; is the upper limit of the water inlet volume flow rate for safe operation of the water pump motor at time t; 、 They are respectively the permitted water-discharge periods and non-water-discharge periods set for water storage facilities during operation and maintenance.

[0025] Furthermore, according to one embodiment of the present invention, the integrated energy system model includes: a converter model, a DC bus power balance model, an AC grid power node model, an AC grid safe operation constraint model, and a DC grid safe operation constraint model; The expression of the converter model is as follows: ; Where: The efficiency of the converter in converting DC to AC; The efficiency of the converter in converting AC to DC; is the DC side input active power when the converter converts DC to AC working mode at time t; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; is the average power factor angle when the converter is working; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; is the tangent function; The expression of the DC bus power balance model is as follows: ; Where: is the net active power injected into the DC distribution network node i at time t; The predicted photovoltaic power generation power at time t; is the power actually consumed by photovoltaic power generation in optimized operation at time t; 、 are the discharge power and charging power of the power storage facility at time t respectively; is the DC side input active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; The expression of the AC power grid node model is as follows: ; Where: is the total active power injected into the AC grid node j at time t; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the data point at time t for predicting uncontrollable power load; To increase the total operating power of the water system at time t; is the total reactive power injected into the AC grid node j at time t; is the reactive power of the node at the tie line of the grid connection point at time t; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; It is the average power factor angle when the uncontrollable power load is working; is the average power factor angle when the water pump motor is working; is the tangent function; The expression of the AC power grid safe operation constraint model is as follows: ; Where: is the voltage amplitude of the AC grid node i at time t; is the maximum upper limit of the voltage amplitude of the AC grid node i; is the minimum lower limit of the voltage amplitude of the AC grid node i; is the voltage amplitude of AC grid node j at time t; is the maximum upper limit of the voltage amplitude at node j in the AC grid; is the minimum lower limit of the voltage amplitude of the AC grid node j; is the AC power grid line at time t Current amplitude; AC grid line The maximum upper limit of the current amplitude; AC grid line Minimum lower limit of current amplitude; is the total active power injected into the AC grid node j at time t; The maximum upper limit of the total active power injected into the AC grid node j, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the total active power injected into the AC grid node j, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the total reactive power injected into the AC grid node j at time t; is the maximum upper limit of the total reactive power injected into the AC grid node j; is the minimum lower limit of the total reactive power injected into the AC grid node j; The expression of the DC grid safe operation constraint model is as follows: ; Where: is the voltage of DC distribution network node i at time t; is the maximum upper limit of the voltage at the DC grid node i; is the minimum lower limit of the voltage at the DC grid node i; is the voltage of DC distribution network node j at time t; is the maximum upper limit of the voltage at node j in the DC grid; is the minimum lower limit of the voltage at node j in the DC grid; is the net active power injected into the DC distribution network node i at time t; The maximum upper limit of the net active power injected into the DC distribution network node i, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the net active power injected into the DC distribution network node i, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the active power flowing from node i to node j in the DC distribution network at time t; is the maximum upper limit of active power flowing from node i to node j in the DC distribution network; is the minimum lower limit of active power flowing from node i to node j in the DC distribution network.

[0026] Further, according to an embodiment of the present invention, the day-ahead economic low-carbon optimized operation strategy includes: an economic optimized operation model, a low-carbon optimized operation model, and an economic low-carbon optimized operation model; The expression of the economic optimization operation model is as follows: ; Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The electricity price purchased from the grid by the integrated energy system of source, grid, load and storage at time t; The electricity price sold by the integrated energy system of source, grid, load and storage to the large power grid at time t; The expression of the low-carbon optimization operation model is as follows: ; Where: To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; is the power actually consumed by photovoltaic power generation in optimized operation at time t; The unit power equivalent carbon emission coefficient of electricity purchased from the large power grid for the integrated energy system of source, grid, load and storage; The expression of the economic low-carbon optimization operation model is as follows: ; Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the economic and low-carbon operation goals of the integrated energy system of source, grid, load and storage during the operation cycle; They are respectively the positive deviation weight of economic optimization target, the negative deviation weight of economic optimization target, the positive deviation weight of low-carbon optimization target, and the negative deviation weight of low-carbon optimization target; They are respectively the positive deviation of economic optimization target, the negative deviation of economic optimization target, the positive deviation of low-carbon optimization target, and the negative deviation of low-carbon optimization target; Acceptable tolerance coefficient for economic optimization objectives; Acceptable tolerance coefficient for low carbon optimization goals; It is the optimal result for single-objective economic optimization; This is the optimal result for single-objective low-carbon optimization.

[0027] Furthermore, according to one embodiment of the present invention, a daily economic low-carbon optimization operation strategy is set for the integrated energy system model, and coupling connection conditions between the daily economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy are set. After the coupling connection conditions are set, an integrated energy system is formed, including: Solve the day-ahead economic low-carbon optimization operation model and obtain the optimal solution. The optimal solution includes: the power value of the grid connection line at time t supplied to the source-grid-load-storage integrated energy system is recorded as The power returned to the large power grid by the source-grid-load-storage integrated energy system at time t is recorded as , the discharge power of the storage facility at time t is recorded as , the charging power of the storage facility at time t is recorded as The total operating power of the water lifting system at time t is recorded as The outflow volume flow rate of the water storage facility at time t is recorded as , when the converter switches from AC to DC working mode at time t, the AC side input active power is recorded as When the converter converts DC to AC working mode at time t, the AC side output active power is recorded as The economic and low-carbon optimization operation target of the source-grid-load-storage integrated energy system under the optimized operation cycle is recorded as ; Set the daily economic low-carbon optimization operation cycle to ; According to the photovoltaic power generation AI prediction model, the daily economic low-carbon optimization operation cycle is re-predicted as The photovoltaic power generation power at time t is recorded as ,use Replace the photovoltaic power generation forecast power at time t in the economic low-carbon optimization operation of the day before ; According to the uncontrollable power load forecasting model, the daily economic low-carbon optimization operation cycle is re-forecasted as The uncontrollable power load power at time t (uncontrollable power load forecast data point) is recorded as ,use Replace the t-time data point of the uncontrollable power load predicted in the day-ahead economic low-carbon optimization operation ; The daily economic low-carbon optimization operation cycle is On this basis, the economic low-carbon optimization operation model is re-solved, and the intraday day-ahead coupled nested boundary conditions (i.e., coupled connection conditions) are additionally set. The expression of the intraday day-ahead coupled nested boundary condition model is as follows: ; Where: To obtain the power value of the grid connection line supplied to the source-grid-load-storage integrated energy system at time t after solving the economic low-carbon optimization operation model of the day before; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The fluctuation tolerance coefficient of power supply power is nested and optimized for intraday and day-ahead coupling; To obtain the total operating power of the water lifting system at time t after solving the day-ahead economic low-carbon optimization operation model; To increase the total operating power of the water system at time t; In order to improve the total power of the water system, the day-ahead coupled nested optimization operation fluctuation tolerance coefficient is adopted; To solve the day-ahead economic low-carbon optimization operation model, the grid connection line is obtained at time t and the power returned to the large power grid by the source-grid-load-storage integrated energy system; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient of the returned power is used; To obtain the AC side input active power when the converter switches from AC to DC working mode at time t after solving the day-ahead economic low-carbon optimization operation model; The active power input on the AC side when the converter is converted from AC to DC working mode; The permissible fluctuation coefficient of the intraday and day-ahead coupled nested optimization operation of the AC side input active power; To solve the economic low-carbon optimization operation model obtained after the day before, the AC side output active power of the converter when the DC is converted to AC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; The permissible fluctuation coefficient of the AC side output active power is determined by nested optimization of intraday and day-ahead coupling operation; To obtain the economic low-carbon optimization operation target value of the source-grid-load-storage integrated energy system under the optimization operation cycle after solving the day-ahead economic low-carbon optimization operation model; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient within the low-carbon optimization operation target day; To obtain the discharge power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the discharge power of the storage facility at time t; The discharge power intraday day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the charging power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the charging power of the power storage facility at time t; The charging power intraday and day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the outflow volume flow of the water storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the outlet volume flow rate of the water storage facility at time t; It is the permissible coefficient of fluctuation of the outlet volume flow rate during the day and the day before the day coupled nested optimization operation.

[0028] Furthermore, according to an embodiment of the present invention, inputting the integrated energy system information includes: Input the integrated energy system information, including the upper limit power of the power supply capacity of the grid-connected point interconnection line to the source-grid-load-storage integrated energy system, the upper limit power of the power return capacity of the grid-connected point interconnection line source-grid-load-storage integrated energy system to the grid, grid connection mode, encoder model parameters, decoder model parameters, comprehensive conversion efficiency of photovoltaic power generation system, effective total surface area of ​​battery components of photovoltaic power generation system, photovoltaic power generation power and temperature regulation conversion coefficient, working environment temperature of solar panels under standard test environment for photovoltaic power generation, weather type influence coefficient, solar radiation intensity, working environment temperature of photovoltaic power generation, relative humidity, wind speed, time label, AC power grid network topology, AC power grid network parameters, AC power grid power Node power generation parameters, pump motor operation mode, day-ahead optimized operation cycle, intraday optimized operation cycle, maximum number of starts within the pump motor operation cycle, pump motor function parameters with respect to water inlet volume flow rate, total number of pump motors in the water lifting system, maximum working time within the pump motor operation cycle, minimum working time within the pump motor operation cycle, operating permission period set within the pump motor operation cycle, historical uncontrollable power load data points, loss rate of the charge and discharge process of the power storage facility, day-ahead optimized operation step resolution, intraday optimized operation step resolution, upper limit of discharge power of the power storage facility, upper limit of charging power of the power storage facility, lower limit of discharge power regulation and protection of the power storage facility, charging power of the power storage facility The lower limit of rate regulation protection, the installed rated capacity of the power storage facility, the functional relationship parameters between the comprehensive energy conversion efficiency of the water pump motor and the water inlet volume flow rate, the upper limit of the water storage capacity of the water storage facility, the lower limit of the water storage capacity of the water storage facility, the lower limit of the water outlet volume flow rate during the permitted water outlet period of the water storage facility, the permitted water outlet time period set for the water storage facility during operation and maintenance, the permitted water outlet time period set for the water storage facility during operation and maintenance, the working efficiency of the converter DC to AC, the working efficiency of the converter AC to DC, the average power factor angle when the converter is working, the average power factor angle when the water pump motor is working, the average power factor angle when the uncontrollable power load is working, the maximum upper limit of the AC grid node voltage amplitude, the maximum AC grid line current amplitude The maximum upper limit, the maximum upper limit of the total reactive power injected into the AC grid nodes, the minimum lower limit of the total reactive power injected into the AC grid nodes, the DC grid network topology, the DC grid network parameters, the maximum upper limit of the DC grid node voltage, the minimum lower limit of the DC grid node voltage, the electricity purchase price, the electricity sales price, the unit power equivalent carbon emission coefficient, the acceptable tolerance coefficient of the economic optimization target, the acceptable tolerance coefficient of the low-carbon optimization target, the positive deviation weight of the economic optimization target, the negative deviation weight of the economic optimization target, the positive deviation weight of the low-carbon optimization target, the negative deviation weight of the low-carbon optimization target, the various intraday day-ahead coupled nested optimization operation fluctuation tolerance coefficients in the intraday day-ahead coupled nested boundary condition model, and other information.

[0029] Furthermore, according to one embodiment of the present invention, outputting the economic and low-carbon optimization result information of the integrated energy system includes: Output integrated energy system information, including the power supplied by the large power grid to the source-grid-load-storage integrated energy system at each moment of the grid connection line, the power returned by the source-grid-load-storage integrated energy system to the large power grid at each moment of the grid connection line, photovoltaic power generation power at each moment, the total active power of each node in the AC grid at each moment, the total reactive power of each node in the AC grid at each moment, the voltage amplitude of each node in the AC grid at each moment, the net active power injected into each node in the DC distribution network at each moment, the voltage of each node in the DC distribution network at each moment, the operating state variables of the water pump motor at each moment, the operating electric power of the water pump motor at each moment, the total operating electric power of the water lifting system at each moment, the predicted value of the uncontrollable power load at each moment, the discharge power of the storage facility at each moment, and the storage facility at each moment. The charging power at each moment, the discharging power of the energy storage facility at each moment, the charging power of the energy storage facility at each moment, the water storage capacity of the water storage facility at each moment, the water outlet volume flow of the water storage facility at each moment, the input active power of the AC side of the converter at each moment, the output active power of the AC side of the converter at each moment, the current amplitude of each line at each moment of the AC power grid, the voltage of each node of the DC distribution network at each moment, the economic optimization operation target of the integrated energy system of source, grid, load and storage under the optimized operation cycle, the low-carbon optimization operation target of the integrated energy system of source, grid, load and storage under the optimized operation cycle, the economic and low-carbon optimization operation target of the integrated energy system of source, grid, load and storage under the optimized operation cycle, the positive deviation of the economic optimization target, the negative deviation of the economic optimization target, the positive deviation of the low-carbon optimization target, the negative deviation of the low-carbon optimization target and other information.

[0030] According to the above scheme of the present invention, the present invention balances economic benefits and emission reduction benefits by optimizing electricity purchase and sales strategies and carbon emission sensitive equipment, enhances the ability of the integrated source-grid-load-storage energy system to cope with the uncertainty of new energy power generation, and fully taps the potential of renewable energy; the present invention improves the operating stability and safety of source-grid-load-storage equipment and AC / DC power grids through comprehensive modeling of source-grid-load-storage and AC / DC power grid safety constraints, optimizes the coordinated regulation and operation strategies between equipment and AC / DC power grids, improves the overall performance of the system, realizes the global optimized operation of the energy system, and provides strong support for the economic and low-carbon operation of the system; the present invention can accurately reflect the actual operation of the energy system through the deep fusion model of AI and mechanism constructed, and provide a more reliable basis for optimized operation. It not only realizes theoretical and method innovation, but also provides a practical and feasible technical solution for the coordinated operation of the integrated source-grid-load-storage energy system under the background of new power system through the engineerable model architecture and solution algorithm.

[0031] Furthermore, to achieve the above objectives, the present invention also provides an energy system economic low-carbon optimization operation system integrating AI and mechanism, comprising: Source-side model building module, which builds the source-side model, including the large-scale power grid power supply model and the photovoltaic power generation AI prediction model; Grid-side model building module, which builds the grid-side model, including the AC grid power flow model and the DC distribution network power balance model; Load-side model building module, which builds the load-side model, including the water pump motor model and the uncontrollable power load prediction model; The reservoir side model establishment module establishes the reservoir side model, including the power storage facility model and the water storage facility model; An integrated energy system model building module couples and interconnects the source side model, grid side model, load side model, and storage side model to form an integrated energy system model; The day-ahead economic low-carbon optimization operation strategy setting module sets the day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; The intraday economic low-carbon optimization operation strategy setting module sets the intraday economic low-carbon optimization operation strategy for the integrated energy system model, and sets the coupling connection conditions between the intraday economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed; The result output module inputs the integrated energy system information into the integrated energy system, and outputs the economic and low-carbon optimization result information of the integrated energy system through the integrated energy system.

[0032] According to the above-mentioned AI and mechanism-integrated energy system economic and low-carbon optimization operation system of the present invention, the above-mentioned AI and mechanism-integrated energy system economic and low-carbon optimization operation method can be realized. The specific process steps are as described above and will not be repeated here.

[0033] Furthermore, to achieve the above-mentioned purpose, the present invention also provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the economic and low-carbon optimization operation method of the energy system integrating AI and mechanism as described above is implemented.

[0034] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the economic and low-carbon optimization operation method of the energy system integrating AI and mechanism as described above is implemented.

[0035] Those skilled in the art will appreciate that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0036] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and equipment can refer to the corresponding processes in the aforementioned method implementation methods and will not be repeated here.

[0037] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0038] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the objectives of the embodiments of the present invention.

[0039] In addition, each functional module in the embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0040] If the functions are implemented as software modules and sold or used as standalone products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the energy-saving signal transmission / reception method according to various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, ROM, RAM, a magnetic disk, or an optical disk.

[0041] The above description is merely a preferred embodiment of the present application and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the inventive concept. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in this application.

[0042] It should be understood that the size of the serial numbers of each step in the content of the invention and the implementation methods of the present invention does not absolutely mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation methods of the present invention.

Claims

1. The energy system economic low-carbon optimization operation method integrating AI and mechanism is characterized by: include: Establish source-side models, including large-scale power grid power supply models and photovoltaic power generation AI prediction models; Establish grid-side models, including AC grid power flow model and DC distribution network power balance model; Establish a load-side model, including a water pump motor model and an uncontrollable power load prediction model; Establish a storage-side model, including power storage facility model and water storage facility model; The source side model, grid side model, load side model and storage side model are coupled and interconnected to form an integrated energy system model; Set up a day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; Setting a daily economic low-carbon optimization operation strategy for the integrated energy system model, and setting coupling connection conditions between the daily economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed. The integrated energy system information is input into the integrated energy system, and the integrated energy system economic and low-carbon optimization result information is output through the integrated energy system.

2. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The expression of the large power grid power supply model is as follows: ； Where: The power supplied by the large power grid to the source-grid-load-storage integrated energy system at the time of the grid connection line; The upper limit power of the power supply capacity of the grid-connected point tie line power grid to the source-grid-load-storage integrated energy system; The grid connection line at the time of the source-grid-load-storage integrated energy system returns power to the large power grid; The power upper limit of the capacity of the power returned to the grid by the source-grid-load-storage integrated energy system of the grid connection line; The photovoltaic power generation AI prediction model is a sequence-to-sequence model architecture based on the long short-term memory network combined with an attention mechanism to process temporal dependency models, including: an encoder part model, a decoder part model, a photovoltaic power generation mechanism model, an input model of the long short-term memory network temporal prediction model, and an output prediction target model of the long short-term memory network temporal prediction model; The expression of the encoder part model is as follows: ； Where: is the output of the forget gate at time t; is the input gate output at time t; is the output gate output at time t; is the candidate cell state at time t; is the current cell state at time t; for Current cell state at any moment; is the hidden state at time t; for Always hide the status; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; 、 、 、 are all weight matrices, corresponding to the weights of the forget gate, input gate, output gate and cell state respectively; 、 、 、 are all bias vectors, corresponding to the bias items of the forget gate, input gate, output gate and cell state respectively; For variables Sigmoid function; For variables The hyperbolic tangent function of The expression of the decoder part model is as follows: ； Where: Hidden state of the decoder at time t, storing decoding process information; For decoder Always hide the state and store the decoding process information; For the previous step The predicted output at the moment is the normalized photovoltaic power value; z is the context vector, which is initially the final hidden state of the encoder; is the attention weight vector, used to calculate the attention score; For variables The hyperbolic tangent function of 、 Both are attention weight matrices, which handle the trainable parameters of the decoder and encoder states respectively; is the hidden state of the encoder at step i; is the unnormalized attention score, which is the correlation strength between the decoder state at time t and the encoder state at time i; is the unnormalized attention score, which is the strength of the association between the decoder state at time t and the encoder state at time j; is the attention weight, the importance of the encoder at moment i to the decoder at moment t; is the dynamic context vector at time t; To predict the output at time t, it is necessary to denormalize to obtain the actual power value; is the output layer weight matrix, mapping the concatenated state to the output; is the output layer bias, which predicts the output bias term; T is the encoder sequence length, that is, the total number of input time steps; It is a complete long short-term memory network unit calculation process function, which contains the input gate that controls the inflow of new information, the forget gate that controls the retention of historical information, the output gate that controls the state output, and the cell state of the long-term memory carrier; The expression of the photovoltaic power generation mechanical model is as follows: ； Where: is the photovoltaic power generation power at time t; The comprehensive conversion efficiency of the photovoltaic power generation system; is the total effective surface area of ​​the photovoltaic power generation system battery components; is the solar radiation intensity at time t; The conversion coefficient between photovoltaic power generation and temperature adjustment; is the working environment temperature of the photovoltaic solar panel at time t; The operating temperature of solar panels under standard test conditions for photovoltaic power generation; is the photovoltaic panel tilt angle correction function; is the tilt angle of the photovoltaic panel; 、 for Different parameters of the function; is the weather type influence coefficient, where Indicates the sunny day coefficient, which is usually set to 1. Indicates the cloudy coefficient, which is generally set to a value between 0.6 and 0.

8. Indicates the rainy day coefficient, which is generally set to a value between 0.2 and 0.5; is the weather type at time t; Convert weather type codes and convert text values ​​at different times into numerical values; The expression of the input model of the long short-term memory network time series prediction model is as follows: ； Where: is the solar radiation intensity at time t; is the working environment temperature of photovoltaic power generation at time t; is the relative humidity at time t; is the wind speed at time t; is the weather type at time t; for Photovoltaic power generation at any moment; for Photovoltaic power generation at any moment; 、 、 、 They are minute, hour, day, and month time labels respectively; The input feature vector at time t contains the standardized PV power generation meteorological data and historical power; The expression of the output prediction target model of the long short-term memory network time series prediction model is as follows: ； Where: It is a mapping function of a complete long short-term memory network neural network to the time series, realizing the input and output mapping function, including the encoder time series and decoder prediction generation process model content; T is the encoder sequence length, that is, the total number of input time steps; n is the total number of historical feature vectors; 、 、 They are time, time, The predicted photovoltaic power generation power at the moment; 、 、 They are time t, time, The input feature vector at the time contains the standardized photovoltaic power generation meteorological data and historical power.

3. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The expression of the AC power grid flow model is as follows: ； Where: is the total active power injected into the AC grid node j at time t; is the total reactive power injected into the AC grid node j at time t; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Active power on top; is the AC power grid line at time t Reactive power; is the AC power grid line at time t Current amplitude; AC grid line resistance; AC grid line reactance; is the conductance of AC grid node j; is the susceptance of AC grid node j; is the voltage amplitude of AC grid node j at time t; is the voltage amplitude of the AC grid node i at time t; is the set of first nodes in the AC power grid with j as the last node, and Indicates that the first node i is connected to the last node j; is the set of tail nodes in the AC power grid with j as the first node, and Indicates that the tail node i is connected to the head node j; The expression of the DC distribution network power balance model is as follows: ； Where: is the net active power injected into the DC distribution network node i at time t; is the active power flowing from node i to node j in the DC distribution network at time t; The power branch formed by the DC distribution network node i and node j at time t Power loss; is the voltage of DC distribution network node i at time t; is the voltage of DC distribution network node j at time t; is the conductance between nodes i and j in the DC distribution network; Node j is directly connected to node i in the DC distribution network.

4. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The water pump motor model includes: a fully autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, a semi-autonomous optimization water pump motor operation mode model, and a set optimization water pump motor operation mode model, and the operation mode of each model is recorded as The operating mode, The operating mode, In actual operation, an operation mode is selected for optimization. The setting of the operation mode meets the following conditions: ； Where: Select a variable for the fully autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it means that the water pump motor model operates using the fully autonomous optimization water pump motor operation mode model. When the value is 0, it means that the water pump motor model does not operate using the fully autonomous optimization water pump motor operation mode model. A variable is selected for the semi-autonomous optimization water pump motor operation mode model, with a value of 0 or 1. When the value is 1, it indicates that the water pump motor model operates using the semi-autonomous optimization water pump motor operation mode model; when the value is 0, it indicates that the water pump motor model does not operate using the semi-autonomous optimization water pump motor operation mode model; The variable is selected for the set-type optimized water pump motor operation mode model, and the value is 0 or 1. When the value is 1, it means that the water pump motor model is operated using the set-type optimized water pump motor operation mode model. When the value is 0, it means that the water pump motor model is not operated using the set-type optimized water pump motor operation mode model. The specific expression of the fully autonomous optimization water pump motor operation mode model is as follows: ； Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The specific expression of the semi-autonomous optimization water pump motor operation mode model is as follows: ； Where: To optimize the operating cycle; For the mth water pump motor The running state variable at the moment, the value is 0 or 1; is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; is the maximum number of starts in the mth water pump motor operation cycle; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; is the maximum working time in the mth water pump motor operation cycle; is the minimum working time in the mth water pump motor operation cycle; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; The specific expression of the set-type optimization water pump motor operation mode model is as follows: ； Where: is the operating state variable of the m-th water pump motor at time t, and its value is 0 or 1; The operation permission period is set in the mth water pump motor operation cycle. During the operation permission period, the water pump motor is allowed to start and run. During the non-permitted period, the water pump motor cannot start and run. The value is 0; is the operating power of the m-th water pump motor at time t; is the functional relationship of the m-th water pump motor with respect to the water inlet volume flow rate; To increase the total operating power of the water system at time t; To increase the total number of pump motors in the water system; The uncontrollable power load prediction model predicts the uncontrollable power load prediction data point results, including: Get 6 uncontrollable power load data points in the recent history, which are recorded as 、 、 、 、 、 ; The next predicted uncontrollable power load data point is recorded as ; Set the moving average calculation step size is 3; Calculate the first moving average. The expression of the first moving average calculation model is as follows: ； Where: is the result of the first moving average calculation at time t; 、 、 They are time t, time, The historical uncontrollable power load data points at the time; The step size for moving average calculation is set; After calculating the first moving average, the data sequence is formed: 、 、 、 ; Calculate the second moving average, and the expression of the calculation model is as follows: ； Where: for The second moving average calculation result of the time; 、 、 They are time t, time, The first moving average calculation result of the moment; The step size for moving average calculation is set; To predict the next latest data point of uncontrollable power load, the prediction model expression is as follows: ， Where: To predict the next latest data point of uncontrollable power load, that is, the predicted The data point at that moment, in particular, when hour, Indicates the first latest data point for predicting future uncontrollable power load; is the result of the first moving average calculation at time t; is the result of the second moving average calculation at time t; A new sequence of uncontrollable power load data points is formed, which is recorded as 、 、 、 、 、 ; Based on the new sequence of uncontrollable power load data points, predict the future , the prediction calculation process is the same as above; According to the need of predicting the future time period, predict the 、 and other data points within the time period, thereby obtaining the results of all data points of uncontrollable power load prediction.

5. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The expression of the electricity storage facility model is as follows: ； Where: 、 are the stored energy of the storage facility at time t and at time t, respectively; is the loss rate of the charging and discharging process of the electricity storage facility; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 are the charging efficiency and discharging efficiency of the power storage facility respectively; To optimize the running step resolution; 、 are the discharge power and charging power of the power storage facility at time t respectively; 、 They are the upper limit of discharge power and charging power of the power storage facility respectively; 、 They are the lower limit of discharge power regulation and protection of the power storage facility and the lower limit of charging power regulation and protection; 、 They are the upper limit coefficient and lower limit coefficient of the real-time storage energy of the power storage facility respectively; The installed rated capacity of the electricity storage facility; 、 Optimize the storage energy at the beginning and end of the operation cycle of the storage facility respectively; In order to optimize the start and end of the operation cycle, the balance coefficient of the energy storage facility at the start and end states is set. The value range is between 0 and 1. When the value is 0, it means that there is no requirement for the balance between the start and end states; when the value is 1, it means that a complete balance is required, and the start and end states are completely consistent. To optimize the operating cycle; The expression of the water storage facility model is as follows: ； Where: The operating power of the water pump motor at time t; is the functional relationship of the pump motor with respect to the water inlet volume flow rate; is the density of water; is the acceleration due to gravity; is the pump motor head at time t, including static head and pipeline loss; is the water inlet volume flow rate of the pump motor at time t; is the comprehensive energy conversion efficiency of the pump motor at time t, including motor and hydraulic efficiency; 、 、 Different parameters for the functional relationship between the pump motor head and the water inlet volume flow rate; 、 、 Different parameters for the functional relationship between the comprehensive efficiency of energy conversion of the water pump motor and the volume flow rate of water inlet; for The water storage capacity of water storage facilities at any given time; is the water storage capacity of the water storage facility at time t; 、 They are the upper and lower limits of water storage capacity of water storage facilities respectively; is the outlet volume flow rate of the water storage facility at time t; The lower limit of the water discharge volume flow rate during the permitted water discharge period of the water storage facility; The maximum upper limit of the water outlet volume flow rate of the water storage facility; is the upper limit of the water inlet volume flow rate for safe operation of the water pump motor at time t; 、 They are respectively the permitted water-discharge periods and non-water-discharge periods set for water storage facilities during operation and maintenance.

6. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The integrated energy system model includes: a converter model, a DC bus power balance model, an AC power grid power node model, an AC power grid safe operation constraint model, and a DC power grid safe operation constraint model; The expression of the converter model is as follows: ； Where: The efficiency of the converter in converting DC to AC; The efficiency of the converter in converting AC to DC; is the DC side input active power when the converter converts DC to AC working mode at time t; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; is the average power factor angle when the converter is working; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; is the tangent function; The expression of the DC bus power balance model is as follows: ； Where: is the net active power injected into the DC distribution network node i at time t; The predicted photovoltaic power generation power at time t; is the power actually consumed by photovoltaic power generation in optimized operation at time t; 、 are the discharge power and charging power of the power storage facility at time t respectively; is the DC side input active power when the converter converts DC to AC working mode at time t; is the DC side output active power when the converter switches from AC to DC working mode at time t; The expression of the AC power grid node model is as follows: ； Where: is the total active power injected into the AC grid node j at time t; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; is the AC side input active power when the converter switches from AC to DC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; is the data point at time t for predicting uncontrollable power load; To increase the total operating power of the water system at time t; is the total reactive power injected into the AC grid node j at time t; is the reactive power of the node at the tie line of the grid connection point at time t; is the reactive power on the AC side when the converter converts DC to AC working mode at time t; is the reactive power on the AC side when the converter switches from AC to DC mode at time t; It is the average power factor angle when the uncontrollable power load is working; is the average power factor angle when the water pump motor is working; is the tangent function; The expression of the AC power grid safe operation constraint model is as follows: ； Where: is the voltage amplitude of the AC grid node i at time t; is the maximum upper limit of the voltage amplitude of the AC grid node i; is the minimum lower limit of the voltage amplitude of the AC grid node i; is the voltage amplitude of AC grid node j at time t; is the maximum upper limit of the voltage amplitude at node j in the AC grid; is the minimum lower limit of the voltage amplitude of the AC grid node j; is the AC power grid line at time t Current amplitude; AC grid line The maximum upper limit of the current amplitude; AC grid line Minimum lower limit of current amplitude; is the total active power injected into the AC grid node j at time t; The maximum upper limit of the total active power injected into the AC grid node j, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the total active power injected into the AC grid node j, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the total reactive power injected into the AC grid node j at time t; is the maximum upper limit of the total reactive power injected into the AC grid node j; is the minimum lower limit of the total reactive power injected into the AC grid node j; The expression of the DC grid safe operation constraint model is as follows: ； Where: is the voltage of DC distribution network node i at time t; is the maximum upper limit of the voltage at the DC grid node i; is the minimum lower limit of the voltage at the DC grid node i; is the voltage of DC distribution network node j at time t; is the maximum upper limit of the voltage at node j in the DC grid; is the minimum lower limit of the voltage at node j in the DC grid; is the net active power injected into the DC distribution network node i at time t; The maximum upper limit of the net active power injected into the DC distribution network node i, and the active power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; The minimum lower limit of the net active power injected into the DC distribution network node i, and the reactive power of each type of power generation and consumption equipment at this point meets its operating physical constraint boundary; is the active power flowing from node i to node j in the DC distribution network at time t; is the maximum upper limit of active power flowing from node i to node j in the DC distribution network; is the minimum lower limit of active power flowing from node i to node j in the DC distribution network.

7. The method for economic low-carbon optimization operation of energy systems integrating AI and mechanism according to claim 1 is characterized in that: The day-ahead economic low-carbon optimization operation strategy includes: an economic optimization operation model, a low-carbon optimization operation model, and an economic low-carbon optimization operation model; The expression of the economic optimization operation model is as follows: ； Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The electricity price purchased from the grid by the integrated energy system of source, grid, load and storage at time t; The electricity price sold by the integrated energy system of source, grid, load and storage to the large power grid at time t; The expression of the low-carbon optimization operation model is as follows: ； Where: To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the operating cycle; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; is the power actually consumed by photovoltaic power generation in optimized operation at time t; The unit power equivalent carbon emission coefficient of electricity purchased from the large power grid for the integrated energy system of source, grid, load and storage; The expression of the economic low-carbon optimization operation model is as follows: ； Where: To optimize the economic operation objectives of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; To optimize the economic and low-carbon operation goals of the integrated energy system of source, grid, load and storage during the operation cycle; 、 、 、 They are respectively the positive deviation weight of economic optimization target, the negative deviation weight of economic optimization target, the positive deviation weight of low-carbon optimization target, and the negative deviation weight of low-carbon optimization target; 、 、 、 They are respectively the positive deviation of economic optimization target, the negative deviation of economic optimization target, the positive deviation of low-carbon optimization target, and the negative deviation of low-carbon optimization target; Acceptable tolerance coefficient for economic optimization objectives; Acceptable tolerance coefficient for low carbon optimization goals; It is the optimal result for single-objective economic optimization; This is the optimal result for single-objective low-carbon optimization.

8. The method for economic low-carbon optimization operation of energy system integrating AI and mechanism according to claim 7 is characterized in that: The integrated energy system model sets a daily economic low-carbon optimization operation strategy, and sets coupling connection conditions between the daily economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed, including: Solve the day-ahead economic low-carbon optimization operation model and obtain the optimal solution. The optimal solution includes: the power value of the grid connection line at the time of the large power grid to the source-grid-load-storage integrated energy system is recorded as The power returned to the large power grid by the source-grid-load-storage integrated energy system at the time of the grid connection line is recorded as , the discharge power of the storage facility at time t is recorded as , the charging power of the storage facility at time t is recorded as The total operating power of the water lifting system at time t is recorded as The outflow volume flow rate of the water storage facility at time t is recorded as , when the converter switches from AC to DC working mode at time t, the AC side input active power is recorded as When the converter converts DC to AC working mode at time t, the AC side output active power is recorded as The economic and low-carbon optimization operation target of the source-grid-load-storage integrated energy system under the optimized operation cycle is recorded as ; Set the daily economic low-carbon optimization operation cycle to ; According to the photovoltaic power generation AI prediction model, the daily economic low-carbon optimization operation cycle is re-predicted as The photovoltaic power generation power at time t is recorded as ,use Replace the photovoltaic power generation forecast power at time t in the economic low-carbon optimization operation of the day before ; According to the uncontrollable power load forecasting model, the daily economic low-carbon optimization operation cycle is re-forecasted as The uncontrollable power load power at time t is recorded as ,use Replace the t-time data point of the uncontrollable power load predicted in the day-ahead economic low-carbon optimization operation ; The daily economic low-carbon optimization operation cycle is On this basis, the economic low-carbon optimization operation model is re-solved, and the intraday day-ahead coupled nested boundary conditions are additionally set. The expression of the intraday day-ahead coupled nested boundary condition model is as follows: ； Where: To obtain the power value of the grid connection line supplied to the source-grid-load-storage integrated energy system at time t after solving the economic low-carbon optimization operation model of the day before; The power supplied by the large power grid to the source-grid-load-storage integrated energy system at time t for the grid connection line; The fluctuation tolerance coefficient of power supply power is nested and optimized for intraday and day-ahead coupling; To obtain the total operating power of the water lifting system at time t after solving the day-ahead economic low-carbon optimization operation model; To increase the total operating power of the water system at time t; In order to improve the total power of the water system, the day-ahead coupled nested optimization operation fluctuation tolerance coefficient is adopted; To solve the day-ahead economic low-carbon optimization operation model, the grid connection line is obtained at time t and the power returned to the large power grid by the source-grid-load-storage integrated energy system; The grid connection line at time t is used to return power to the large power grid through the integrated energy system of source, grid, load and storage; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient of the returned power is used; To obtain the AC side input active power when the converter switches from AC to DC working mode at time t after solving the day-ahead economic low-carbon optimization operation model; is the AC side input active power when the converter switches from AC to DC working mode at time t; The permissible fluctuation coefficient of the intraday and day-ahead coupled nested optimization operation of the AC side input active power; To solve the economic low-carbon optimization operation model obtained after the day before, the AC side output active power of the converter when the DC is converted to AC working mode at time t; is the AC side output active power when the converter converts DC to AC working mode at time t; The permissible fluctuation coefficient of the AC side output active power is determined by nested optimization of intraday and day-ahead coupling operation; To obtain the economic low-carbon optimization operation target value of the source-grid-load-storage integrated energy system under the optimization operation cycle after solving the day-ahead economic low-carbon optimization operation model; To optimize the low-carbon operation goals of the source-grid-load-storage integrated energy system during the operation cycle; The day-ahead coupled nested optimization operation fluctuation tolerance coefficient within the low-carbon optimization operation target day; To obtain the discharge power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the discharge power of the energy storage facility at time t; The discharge power intraday day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the charging power of the power storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the charging power of the power storage facility at time t; The charging power intraday and day-ahead coupled nested optimization operation fluctuation tolerance coefficient; To obtain the outflow volume flow of the water storage facility at time t after solving the day-ahead economic low-carbon optimization operation model; is the outlet volume flow rate of the water storage facility at time t; It is the permissible coefficient of fluctuation of the outlet volume flow rate during the day and the day before the day coupled nested optimization operation.

9. The energy system economic low-carbon optimization operation system integrating AI and mechanism is characterized by: include: Source-side model building module, which builds the source-side model, including the large-scale power grid power supply model and the photovoltaic power generation AI prediction model; Grid-side model building module, which builds the grid-side model, including the AC grid power flow model and the DC distribution network power balance model; Load-side model building module, which builds the load-side model, including the water pump motor model and the uncontrollable power load prediction model; The reservoir side model establishment module establishes the reservoir side model, including the power storage facility model and the water storage facility model; An integrated energy system model building module couples and interconnects the source side model, grid side model, load side model, and storage side model to form an integrated energy system model; The day-ahead economic low-carbon optimization operation strategy setting module sets the day-ahead economic low-carbon optimization operation strategy for the integrated energy system model; The intraday economic low-carbon optimization operation strategy setting module sets the intraday economic low-carbon optimization operation strategy for the integrated energy system model, and sets the coupling connection conditions between the intraday economic low-carbon optimization operation strategy and the day-ahead economic low-carbon optimization operation strategy. After the coupling connection conditions are set, an integrated energy system is formed; The result output module inputs the integrated energy system information into the integrated energy system, and outputs the economic and low-carbon optimization result information of the integrated energy system through the integrated energy system.

10. An electronic device, characterized in that The invention comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the method for economic and low-carbon optimization operation of an energy system integrating AI and mechanism as described in any one of claims 1 to 8 is implemented.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the energy system economic low-carbon optimization operation method integrating AI and mechanism according to any one of claims 1 to 8.

Images