Multi-supply cooperative scheduling system and method for regional integrated energy system

Abstract

The invention relates to the technical field of energy coordination, in particular to a multi-supply cooperative scheduling system and method for a regional integrated energy system, and the system comprises a data collection module which is used for obtaining an equipment operation state data set and an energy load demand data set in the regional integrated energy system in real time; the multi-element collaborative strategy generation module is used for calculating the equipment operation state data set and the energy consumption load demand data set by using a collaborative scheduling model based on a multi-element main body game to obtain a multi-element collaborative scheduling strategy; the strategy updating module is used for acquiring external environment information in real time and inputting the external environment information and the multivariate collaborative scheduling strategy into a preset strategy updating node analysis model to generate a next strategy updating time node; and the duration feature extraction module is used for collecting a plurality of continuous strategy update time nodes and performing duration feature extraction. According to the method, collaborative optimization scheduling of multiple energy forms can be realized in a complex and changeable environment, and the energy utilization efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy coordination, and particularly to a multi-source supply collaborative scheduling system and method for a regional integrated energy system. Background Art

[0002] With the continuous growth of global energy demand and the transformation of the energy structure, the regional integrated energy system, as a complex energy supply and management network integrating various energy forms (such as electricity, heat, cold, gas, etc.), various energy equipment, and various energy users, has received increasing attention. In the regional integrated energy system, different types of energy supply equipment, such as power generation equipment, heating equipment, gas supply equipment, etc., need to operate collaboratively during operation to meet the energy demand of various energy-consuming load groups in the region.

[0003] The operation of the regional integrated energy system is not only affected by the operation status of its own equipment and the energy load demand, but also restricted by the external environment, such as grid electricity price, natural gas price, and meteorological data. External environmental factors are constantly changing, which will affect the supply cost and demand characteristics of energy, and further require the scheduling strategy to be dynamically adjusted according to the changes in the external environment. Most of the existing scheduling strategies for regional integrated energy systems adopt a fixed update cycle setting, lacking quantitative analysis of the characteristics of update nodes, resulting in the model being difficult to adapt to the complex and changeable internal and external environment. Especially when the external environment fluctuates violently, the fixed update cycle is likely to cause strategy lag or frequent adjustment. Summary of the Invention

[0004] The present invention provides a multi-source supply collaborative scheduling system and method for a regional integrated energy system, which can realize the collaborative optimal scheduling of various energy forms in a complex and changeable environment, improve energy utilization efficiency, and can effectively solve the problems in the background art.

[0005] To achieve the above object, in a first aspect, the present invention provides a multi-source supply collaborative scheduling system for a regional integrated energy system, including:

[0006] A data acquisition module for real-time obtaining a device operation status data set and an energy load demand data set in the regional integrated energy system;

[0007] A multi-source collaborative strategy generation module for operating the device operation status data set and the energy load demand data set by using a built-in multi-source collaborative scheduling model to obtain a multi-source collaborative scheduling strategy;

[0008] A strategy update module for real-time obtaining external environment information and inputting it together with the multi-source collaborative scheduling strategy into a preset strategy update node analysis model to generate the next strategy update time node;

[0009] The duration feature extraction module is used to collect multiple consecutive policy update time nodes, perform duration feature extraction, and obtain multiple policy durations;

[0010] The stability evaluation module is used to eliminate external environmental factors for each of the policy durations, and perform stability evaluation on the multiple policy durations after elimination to obtain a stability index;

[0011] The model optimization and deployment module is used to optimize the multi - collaborative scheduling model according to the difference between the stability index and a preset stability threshold, and update and deploy the optimized model to the regional integrated energy system.

[0012] Combined with the first aspect, in a possible design, the duration feature extraction module is further configured to:

[0013] The multiple consecutive policy update time nodes at least include the current policy update time node, the next policy update time node, and the historical policy update time node that is closest to the current policy update time node in the time dimension.

[0014] In a second aspect, the present invention also provides a multi - supply collaborative scheduling method for a regional integrated energy system, including:

[0015] Obtain the device operation status data set and the energy consumption load demand data set in the regional integrated energy system in real - time, and input both into the multi - collaborative scheduling model to obtain a multi - collaborative scheduling policy;

[0016] Obtain external environmental information in real - time, and input it and the multi - collaborative scheduling policy into the policy update node analysis model to generate the next policy update time node;

[0017] Collect multiple consecutive policy update time nodes, and perform duration feature extraction to obtain multiple policy durations;

[0018] Eliminate external environmental factors for each of the policy durations, and perform stability evaluation on the multiple policy durations after elimination to obtain a stability index;

[0019] Optimize the multi - collaborative scheduling model based on the difference between the stability index and a preset stability threshold, and update and deploy the optimized multi - collaborative scheduling model to the regional integrated energy system.

[0020] Combined with the second aspect, in a possible design, the external environmental information includes grid electricity price, natural gas price, and meteorological data.

[0021] Combined with the second aspect, in a possible design, the device operating status data set includes the operating status information of various energy supply devices, and the energy consumption load demand data set includes the real-time energy consumption demand information of different types of energy consumption load groups.

[0022] Combined with the second aspect, in a possible design, the multiple consecutive policy update time nodes at least include the current policy update time node, the next policy update time node, and the historical policy update time node that is closest to the current policy update time node in the time dimension.

[0023] Combined with the second aspect, in a possible design, the policy duration represents the time span between two adjacent policy update time nodes;

[0024] The multiple policy durations at least include the policy duration before the current update and the policy duration after the current update.

[0025] Combined with the second aspect, in a possible design, the multi-source collaborative scheduling model adopts a composite model, including a supply-demand interaction game model, a user group evolutionary game model, and a benefit evaluation model.

[0026] Combined with the second aspect, in a possible design, the active update trigger condition of the policy update node analysis model is: when the comprehensive fitness of the current multi-source collaborative scheduling policy decays to a set fitness threshold, it is determined that the current multi-source collaborative scheduling policy fails and needs to be actively updated;

[0027] The formula for calculating the comprehensive fitness is:

[0028] ;

[0029] where F(t) represents the comprehensive fitness; t represents the duration of the current multi-source collaborative scheduling policy; F0 represents the normalized initial fitness, indicating that the initial state of the policy is optimal; -ΔC price ·t represents the linear decay of the price shock over time. The greater the shock, the faster the comprehensive fitness decreases; -ΔE weather ·t 2 represents the quadratic decay of the meteorological shock over time, reflecting the non-linear impact of emergencies; A s +A d represents the weighted sum of the adaptability between the supply side and the demand side. The larger the value, the more robust the policy;

[0030] By inversely solving the duration of the current multi-source collaborative scheduling policy through the set fitness threshold and combining the update time node of the current policy, the next update time node is calculated.

[0031] Combined with the second aspect, in a possible design, the ΔC price represents the impact factor of price fluctuation, and its calculation formula is:

[0032] ;

[0033] where P e (t) represents the electricity price at time t; P e (t - Δt) represents the electricity price at the previous monitoring time point (t - Δt); P g (t) represents the natural gas price at the current time t, with the unit of yuan / m³; P g (t - Δt) represents the natural gas price at the previous monitoring time point (t - Δt); d1 max represents the upper limit of the electricity price peak in the floating range of time-of-use electricity price on the demand side; d1 min represents the lower limit of the electricity price valley in the floating range of time-of-use electricity price on the demand side; P g base represents the historical benchmark value of the natural gas price, which is used to measure the degree to which the current gas price deviates from the historical average.

[0034] Through the technical solution of the present invention, the following technical effects can be achieved: Through multi-module closed-loop linkage, an adaptive mechanism of dynamic perception, strategy generation, node optimization, and stability iteration is constructed, breaking through the limitations of static update and single-objective optimization of traditional scheduling systems, and achieving efficient, stable, and sustainable collaborative scheduling in complex and changeable energy supply and demand scenarios; The data acquisition module and the multi-source collaborative strategy generation module cooperate to generate an initial strategy based on real-time and accurate equipment and load data, laying a foundation for the operation of the system; The strategy update module determines a new update time node in combination with external environment information, making the scheduling no longer rely on a fixed cycle and being able to flexibly adapt to environmental changes; The duration feature extraction module collects the durations of multiple update nodes, and the stability evaluation module eliminates the external environmental factors to evaluate the stability on this basis, which can accurately reflect the inherent stability characteristics of the system strategy adjustment. Finally, the model optimization and deployment module optimizes the model based on the stability index and updates the deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 is the structural block diagram of the multi-source supply collaborative scheduling system for the regional integrated energy system in the present invention;

[0037] Figure 2 This is the logic flowchart of the multi - supply collaborative scheduling system for the regional integrated energy system in the present invention;

[0038] Figure 3 This is the logic flowchart of the supply - demand interaction game model, user group evolutionary game model and benefit evaluation model in the present invention. Detailed implementation manners

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0040] Next, the present application will be described in conjunction with the accompanying drawings in the present application.

[0041] As Figures 1 to 2 shown, a multi - supply collaborative scheduling system for the regional integrated energy system of the present invention specifically includes: a data acquisition module, a multi - collaborative strategy generation module, a strategy update module, a duration feature extraction module, a stability evaluation module, and a model optimization and deployment module; the specific implementation is as follows:

[0042] The data acquisition module obtains the device operation status data set and the energy consumption load demand data set in the regional integrated energy system in real time; wherein, the device operation status data set includes the operation status information of various energy supply devices, and the energy consumption load demand data set includes the real - time energy consumption demand information of different types of energy - consuming load groups; the device operation status data set and the energy consumption load demand data set correspond one - to - one in the space - time dimension;

[0043] The multi - collaborative strategy generation module uses the built - in multi - collaborative scheduling model to operate on the device operation status data set and the energy consumption load demand data set to obtain a multi - collaborative scheduling strategy; wherein, the multi - collaborative scheduling model adopts a composite model, including a supply - demand interaction game model, a user group evolutionary game model, and a benefit evaluation model;

[0044] The strategy update module obtains external environment information in real time, and inputs it and the multi - collaborative scheduling strategy into a preset strategy update node analysis model to generate the next strategy update time node; the external environment information includes grid electricity price, natural gas price, and meteorological data;

[0045] The duration feature extraction module collects multiple consecutive policy update time nodes, performs duration feature extraction, and obtains multiple policy durations; the multiple consecutive policy update time nodes at least include the current policy update time node, the next policy update time node, and the historical policy update time node closest to the current policy update time node in the time dimension; the policy duration represents the time span between two adjacent policy update time nodes; the multiple policy durations at least include the policy duration before the current update and the policy duration after the current update;

[0046] The stability evaluation module eliminates external environmental factors for each of the policy durations, and performs stability evaluation on the multiple policy durations after elimination to obtain a stability index;

[0047] The model optimization and deployment module optimizes the multi - collaborative scheduling model according to the difference between the stability index and a preset stability threshold, and updates and deploys the optimized model to the regional integrated energy system.

[0048] In this embodiment, through the closed - loop linkage of multiple modules, an adaptive mechanism of dynamic perception, policy generation, node optimization, and stability iteration is constructed, breaking through the limitations of the traditional scheduling system's static update and single - objective optimization, and realizing efficient, stable, and sustainable collaborative scheduling in complex and changeable energy supply - demand scenarios; the policy update module adjusts the update time node in real time based on external environmental fluctuations (such as electricity price, meteorology), while the stability evaluation module accurately identifies the system's own stability defects by eliminating external environmental factor interference; after the two are combined, the system can not only quickly respond to external mutations (such as extreme weather), but also avoid frequent policy adjustments caused by environmental noise, achieving a balance between sensitive response and operating stability; for example, when the electricity price fluctuates violently, the system can shorten the update cycle to reduce costs, and at the same time suppress excessive optimization caused by short - term fluctuations through stability evaluation, preventing frequent start - stop of equipment;

[0049] The stability evaluation module extracts pure internal operation features (such as equipment aging, load evolution law) by eliminating external environmental factors, enabling the model optimization and deployment module to specifically strengthen the system's endogenous vulnerable links; for example, after eliminating the influence of meteorology on the heating load, if the stability index is still lower than the threshold, the problem of insufficient equipment coordination efficiency is directly exposed, and the optimization algorithm can preferentially adjust the weight parameters of the supply - demand game model instead of blindly increasing the proportion of renewable energy;

[0050] The duration feature extraction module analyzes the historical patterns of the strategy's continuous period (such as seasonality and event-driven features), combines them with the real-time generated update time nodes, enabling the system to possess the dual characteristics of short-term agility and long-term robustness. For example, during the winter heating season, the system can autonomously learn the strong correlation between meteorology and load, automatically extend the strategy update period to match the thermal inertia feature, while retaining the emergency response ability to sudden cold snaps. The above multi-time scale coupling mechanism enables the system to exhibit human-like decision-making adaptability in complex environments.

[0051] In the multi-source supply collaborative scheduling system of the regional integrated energy system, the adaptive mechanisms of dynamic perception, strategy generation, node optimization, and stability iteration, through the closed-loop linkage of multiple modules, achieve effective response to complex and changing energy supply and demand scenarios, specifically as follows:

[0052] The data acquisition module obtains the equipment operation status dataset and the energy consumption load demand dataset in the regional integrated energy system in real time. At the same time, the strategy update module obtains external environment information such as grid electricity prices, natural gas prices, and meteorological data in real time, grasps the internal and external situations of the system, and provides accurate basis for subsequent decision-making.

[0053] The multi-source collaborative strategy generation module uses the built-in multi-source collaborative scheduling model to calculate the equipment operation status and energy consumption load demand data. This model adopts a composite model, including a supply-demand interaction game model, a user group evolution game model, and a benefit evaluation model, comprehensively considers multiple factors, generates a multi-source collaborative scheduling strategy, and realizes the reasonable allocation of energy.

[0054] The strategy update module inputs the external environment information and the multi-source collaborative scheduling strategy into the preset strategy update node analysis model to calculate indicators such as the comprehensive fitness. When the comprehensive fitness decays to the set threshold or the external shock factor exceeds the preset value, the next strategy update time node is determined, enabling the scheduling to flexibly adapt to environmental changes and avoid strategy lag.

[0055] The duration feature extraction module collects the strategy duration obtained at the strategy update time node, and the stability evaluation module evaluates the stability after eliminating external environmental factors to obtain the stability index. The model optimization and deployment module optimizes and updates the deployment of the multi-source collaborative scheduling model according to the difference between this index and the preset threshold, continuously improves the system stability and energy utilization efficiency, and realizes sustainable collaborative scheduling.

[0056] The present invention also provides a multi-source supply collaborative scheduling method for a regional integrated energy system, specifically including the following steps:

[0057] Step S1: Obtain the device operation status data set and the energy consumption load demand data set in the regional integrated energy system in real time, and input both into the multi - collaborative scheduling model to obtain the multi - collaborative scheduling strategy; the device operation status data set and the energy consumption load demand data set correspond one - to - one in the space - time dimension;

[0058] Step S2: Obtain the external environment information in real time, and input it and the multi - collaborative scheduling strategy into the policy update node analysis model to generate the next policy update time node;

[0059] Step S3: Collect multiple consecutive policy update time nodes, and perform duration feature extraction to obtain multiple policy duration times;

[0060] Step S4: Eliminate the external environmental factors for each of the policy duration times, and perform stability evaluation on the multiple policy duration times after elimination to obtain the stability index;

[0061] Step S5: Optimize the multi - collaborative scheduling model based on the difference between the stability index and the preset stability threshold, and update and deploy the optimized multi - collaborative scheduling model to the regional integrated energy system.

[0062] In this embodiment, Step S1 ensures that the basic data for scheduling decisions is always up - to - date and accurate by obtaining the device operation status and the energy consumption load demand data set in real time and inputting them into the multi - collaborative scheduling model to generate a strategy. Combining with the real - time monitoring and analysis of external environment information in Step S2, it can quickly respond to external changes, such as electricity price fluctuations or weather changes, so as to realize the immediate adjustment of the scheduling strategy and avoid the lag problem of the fixed - cycle scheduling strategy. Steps S3 and S4 eliminate the influence of external environmental factors by performing duration feature extraction and stability evaluation on multiple consecutive policy update time nodes to obtain the stability index, which not only improves the ability to cope with complex and changeable internal and external environments, but also reduces the phenomenon of frequent policy adjustments caused by drastic fluctuations in the external environment, enhancing the overall stability and reliability. Step S5 optimizes and redeploys the multi - collaborative scheduling model according to the stability evaluation results to ensure that it is always in the optimal configuration state. Through closed - loop feedback control, it realizes the whole - process management from data collection, strategy generation, external environment analysis, stability evaluation to model optimization, forming a complete dynamic regulation cycle, which not only improves energy utilization efficiency, reduces operation costs, but also realizes the efficient integration and coordinated operation of various energy forms and devices in the region.

[0063] In some embodiments of the present invention, a dynamic multi - collaborative scheduling strategy is generated through real - time data collection and composite model calculation. The data collection is as follows:

[0064] For the equipment operation status dataset, the operation parameters of energy supply equipment such as electricity, heat, cold, and gas are collected in real time through Internet of Things sensors, including but not limited to the output of power generation units, the heating efficiency of heat pumps, the pressure of gas storage tanks, the coefficient of performance of refrigeration equipment, etc. At the same time, discrete events such as the start-stop status and fault alarms of the equipment are recorded; the data is marked in space and time according to a unified timestamp and geographical coordinates, such as GIS grids or equipment codes;

[0065] For the energy consumption load demand dataset, through terminal devices such as smart meters and heat meters, the real-time demand baselines of electricity, heat, cold, and gas are collected by classifying user groups such as industrial, commercial, and residential users, and meteorological compensation parameters such as temperature and humidity are introduced to normalize the data; for example, when the temperature drops suddenly, the heating demand will increase significantly, and such correlations need to be dynamically corrected at the data level; all load data needs to be strictly aligned with the equipment status data in the space-time dimension to ensure that the cooling load demand of a building complex within the same time slice can be accurately mapped to the operation parameters of the corresponding refrigeration unit.

[0066] Furthermore, as Figure 3 shown, the multi-source collaborative scheduling model adopts a composite model, which is jointly composed of a supply-demand interaction game model, a user group evolutionary game model, and a benefit evaluation model, and forms a closed loop through a data flow series connection and feedback mechanism. The specific implementation is as follows:

[0067] The supply-demand interaction game model is based on the equipment operation status and load demand data, and simulates the game behavior between the energy supply side (power generation, heating, gas equipment, and waste heat recovery systems) and the demand side (user groups). The supply-side game is constrained by the equipment output cost and carbon emissions, and solves the Nash equilibrium solution of the coordinated output of multi-energy equipment. For example, when a gas turbine and a heat pump meet the electricity and heat demands, the distributed optimization algorithm is used to dynamically allocate the output ratio; the demand-side game introduces real-time electricity price and heat price signals to simulate the elastic response of user loads to price changes. For example, when the electricity price rises, users may reduce non-essential electricity consumption, and the system needs to recalculate the supply-demand balance point; the input of the supply-demand interaction game model is the equipment output capacity parameters in the equipment operation status data and the real-time demand baseline in the energy consumption load demand data; the output is the preliminary scheduling strategy, including the equipment output plan and the demand-side price signal.

[0068] The user group evolutionary game model is used to analyze the dynamic evolution path of the energy consumption behavior of the user group, including the user strategy revenue function, which defines weight parameters such as economic cost and comfort preference. For example, the cost sensitivity weight of industrial users is set to 0.7, and the comfort weight of residential users is set to 0.6; the replicator dynamics equation is used to describe the diffusion of user strategies, and the Monte Carlo method is combined to predict the load change trend; for example, in high-temperature weather, the residential air-conditioning load may increase in an exponential curve; the input of the user group evolutionary game model is the real-time demand of users in the energy consumption load demand data and the dynamic price signal output by the supply-demand game model; the output is the corrected load demand prediction and suggestions for the evolution of user behavior.

[0069] The core mechanism of the benefit evaluation model is to comprehensively evaluate the benefits of the scheduling strategy in multiple dimensions, including:

[0070] Economic index: Calculate the equipment operation cost and energy procurement cost. For example, the cost per megawatt-hour of output of a gas turbine is 200 yuan;

[0071] Environmental protection index: Evaluate the carbon emission intensity. For example, the carbon emission per megawatt-hour of a coal-fired unit is 0.8 tons;

[0072] Reliability index: Analyze the probability of power supply interruption. For example, the energy storage system needs to ensure 99.9% power supply availability;

[0073] User satisfaction index: Based on parameters such as load satisfaction rate and price acceptance, for example, the user satisfaction threshold is set to 90%;

[0074] Its input is the equipment output plan output by the supply-demand game model and the load demand prediction corrected by the user evolution model; the output is the multi-source collaborative scheduling strategy with the highest comprehensive score.

[0075] The above dynamic calculation process is briefly described as follows: The supply-demand interaction game model generates the initial strategies of equipment output and price incentives; the user evolution model predicts the load change based on the price signal and feeds it back to the game model for iterative optimization; the benefit evaluation model screens the Pareto optimal solution to determine the final scheduling strategy.

[0076] The multi-source collaborative scheduling strategy includes supply-side instructions and demand-side instructions. The supply-side instructions include the output information of various energy equipment and the switching temperature threshold of the waste heat recovery system; the demand-side instructions include user response instruction information, time-of-use electricity price floating range, and interruptible load priority list.

[0077] In this embodiment, the device output capacity and user demand are integrated in real time through a supply-demand interaction game model, and the supply-demand balance is adjusted elastically in combination with price signals; the user group evolutionary game model quantifies economic costs and comfort preferences, and uses the Monte Carlo method to simulate the sharp increase in air-conditioning load under high-temperature weather, improving the demand prediction accuracy and reducing the supply-demand deviation; the benefit evaluation model integrates multi-dimensional indicators such as economy, environmental protection, and reliability, and screens non-inferior solutions through the Pareto frontier; the three sub-models form a closed loop of game, evolution, and evaluation to achieve continuous optimization; after the demand-side electricity price is adjusted, the evolutionary model feeds back the load change and drives the game model to recalculate the device output; the supply-side command coordinates the multi-energy complementarity of electricity, heat, and gas. For example, the waste heat recovery system automatically shuts down during the low valley of heating demand to reduce redundant losses; the demand-side time-of-use electricity price guides users to use energy during off-peak hours to improve the overall energy efficiency.

[0078] In some embodiments of the present invention, in order to accurately predict the strategy update nodes of the regional integrated energy system, the external environment information includes grid electricity price, natural gas price, and meteorological data; the data sources and acquisition methods of the external environment information include:

[0079] Grid electricity price data: The current grid time-of-use electricity price, such as peak-valley-flat electricity price, is obtained in real time through the power trading system, and the electricity price volatility is recorded, such as the maximum / minimum price difference per hour.

[0080] Natural gas price data: Access the natural gas trading market database to obtain the real-time price of regional pipeline natural gas and the predicted price for the next 24 hours.

[0081] Meteorological data: Real-time data such as temperature, humidity, and wind speed in the region are obtained from the meteorological website, and combined with the weather forecast information for the next 6 hours, such as temperature mutation and rainfall probability.

[0082] Before using the above external environment information, data preprocessing and feature extraction are also required, specifically including: normalizing the grid electricity price data, calculating the deviation degree of the current electricity price from the historical average value, such as Z-Score standardization, to quantify the electricity price fluctuation intensity; extracting short-term trend features from the natural gas price data, such as the slope of the linear regression fitting of the future price change; detecting mutations in the temperature in the meteorological data. For example, if the temperature change rate within the sliding window exceeds 3°C / hour, it is determined as a mutation event.

[0083] In this embodiment, the multi-source fusion and dynamic analysis of grid electricity price, natural gas price, and meteorological data can perceive the core external factors affecting energy supply and demand in real time; for example, when high-temperature weather causes a sharp increase in air-conditioning load and the grid electricity price is at a peak, the strategy update node analysis model will immediately shorten the strategy update time node, quickly respond to the load change, and switch to low-cost energy.

[0084] More specifically, the calculation process of the policy update node analysis model includes:

[0085] First, calculate the impact factor of price fluctuation, and the calculation formula is as follows:

[0086] ;

[0087] Among them, ΔC price represents the impact factor of price fluctuation;

[0088] P e (t) represents the electricity price at the current time t, with the unit of yuan / kWh; P e (t - Δt): the electricity price at the previous monitoring time point (t - Δt); calculate the absolute value of the fluctuation of the electricity price within the Δt time window; for example, if the current electricity price is 1.5 yuan / kWh and the previous moment is 1.2 yuan / kWh, the fluctuation range is 0.3 yuan / kWh; P g (t): the natural gas price at the current time t, with the unit of yuan / m³; P g (t - Δt): the natural gas price at the previous monitoring time point; reflect the absolute value of the short-term fluctuation of the natural gas price; for example, if the current gas price is 3.0 yuan / m³ and the previous moment is 2.8 yuan / m³, the fluctuation range is 0.2 yuan / m³;

[0089] d1 max represents the upper limit of the electricity price peak defined in the time-of-use electricity price floating range on the demand side; for example, the upper limit of the peak electricity price is 1.8 yuan / kWh; d1 min represents the lower limit of the electricity price valley defined in the time-of-use electricity price floating range on the demand side; for example, the lower limit of the valley electricity price is 0.8 yuan / kWh; the denominator is the preset floating range of the electricity price (1.8 - 0.8 = 1.0 yuan / kWh), which is used to standardize the electricity price fluctuation; if the actual absolute value of the fluctuation (the numerator) exceeds the denominator range, the volatility > 1, indicating that the electricity price exceeds the allowed fluctuation range and emergency intervention is required;

[0090] P g base represents the historical benchmark value of the natural gas price, usually taking the average price in a certain past period; for example, if the average gas price in the past 30 days is 2.5 yuan / m³, then P g base = 2.5, which is used to measure the degree of deviation of the current gas price from the historical average; if the gas price volatility > 1, such as the current gas price of 5.0 yuan / m³ and the benchmark price of 2.5 yuan / m³, the volatility is 1.0, indicating that the gas price deviates significantly from the normal level.

[0091] Secondly, calculate the impact factor of meteorological mutation, and the calculation formula is as follows:

[0092] ;

[0093] Among them, ΔE weather represents the impact factor of meteorological mutation;

[0094] T(t) - T(t - Δt) represents the absolute value of the temperature difference between the current temperature and the previous time period; for example, when the temperature suddenly rises from 25°C to 30°C, the difference is 5°C;

[0095] s3 represents the switching temperature threshold of the waste heat recovery system; for example, if the waste heat recovery system starts when the temperature is higher than 80°C, then s3 = 80°C; the denominator is the threshold, and when the temperature difference approaches the threshold, the meteorological impact index will be significantly amplified;

[0096] H(t) - H(t - Δt) represents the absolute value of the humidity difference between the current humidity and the previous time period; for example, when the humidity drops from 60% to 50%, the difference is 10%;

[0097] α T represents the temperature impact weight, reflecting the degree of influence of temperature change on the system load; for example, in the heating-dominated area, the temperature weight is relatively high; α H represents the humidity impact weight, which is applicable to scenarios sensitive to humidity, such as data center cooling.

[0098] Then, calculate the adaptation factor of the supply side to the multi-source collaborative scheduling strategy, and the calculation formula is as follows:

[0099] ;

[0100] Among them, A s represents the adaptation factor of the supply side to the multi-source collaborative scheduling strategy;

[0101] N s represents the number of various energy equipment;

[0102] s i current represents the current output value of the equipment; for example, the current charge and discharge power of the energy storage system is 50kW;

[0103] s i max represents the maximum output capacity of the equipment; for example, the rated power of the energy storage system is 100kW;

[0104] s i flex = s i max - s i currentIndicates the adjustable margin of the device; for example, the remaining adjustable power of the energy storage system is 50 kW; if the current output of the device is close to the upper limit, the score of this item is relatively low, indicating insufficient adjustment ability of the device and the need to update the strategy as soon as possible; the formula takes the average value of multiple devices to measure the overall flexibility of the supply side.

[0105] After that, calculate the adaptation factor of the demand side to the multi-source collaborative scheduling strategy, and the calculation formula is as follows:

[0106] ;

[0107] Among them, A d Indicates the adaptation factor of the demand side to the multi-source collaborative scheduling strategy;

[0108] N d Indicates the number of instructions in the user response instruction information;

[0109] d j actual Indicates the actual user response effect; for example, during the peak period of time-of-use electricity price, the user actually reduces the load by 15%;

[0110] d j target Indicates the instruction target value; for example, the demand side instruction requires a 20% load reduction during the peak period;

[0111] w j Indicates the priority weight of the jth instruction, from the list of interruptible load priorities in the demand side instruction; for example, the weight of industrial users is 0.8 and that of commercial users is 0.5;

[0112] If the actual response rate is lower than the target and the priority is high, it indicates that the execution effect of the demand side is poor; the weighted sum of multiple instructions (Nd is the number of instructions) is used to comprehensively evaluate the effectiveness of the demand side strategy.

[0113] Immediately afterwards, calculate the comprehensive fitness, and the calculation formula is as follows:

[0114] ;

[0115] Among them, F(t) represents the comprehensive fitness;

[0116] t represents the duration of the strategy;

[0117] F0 represents the normalized initial fitness, indicating that the initial state of the strategy is optimal;

[0118] -ΔC price ·t represents that the price impact decays linearly with time, and the greater the impact, the faster the fitness decreases;

[0119] -ΔE weather ·t 2It indicates that the meteorological impact decays with the square of time, reflecting the non-linear impact of emergencies; for example, when high temperature persists, the cumulative effect of meteorological impact accelerates the decline of fitness.

[0120] A s +A d It represents the weighted sum of the adaptabilities on the supply side and the demand side. The larger the value, the more robust the strategy indicates.

[0121] Finally, set the active update trigger condition. When the comprehensive fitness decays to the set fitness threshold, it is determined that the current strategy fails and needs to be actively updated. For example, set the fitness threshold to 70%. If ΔC price =0.3, ΔE weather =5, A s +A d =0.8, then:

[0122] 0.7 = 1·e -0.3t-5t^2 ·0.8;

[0123] Solve the above formula to obtain t≈0.4 hours; that is, the sustainable duration of the current strategy is 0.4 hours, and the active strategy update is carried out after 0.4 hours.

[0124] On the other hand, passive update trigger conditions can also be set. For example, if any item of the impact factor of price fluctuation and the impact factor of meteorological mutation exceeds its corresponding preset trigger value, forced update is immediately carried out.

[0125] In this embodiment, the strategy update node analysis model realizes the transformation of the abstract system state into computable mathematical indicators through hierarchical calculation, dynamic decay, and multi-source fusion; based on time decay and threshold trigger, it balances stability and real-time performance; coordinates the equipment capabilities on the supply side and the user behaviors on the demand side to improve the overall energy efficiency; is applicable to scenarios such as electricity market trading, microgrid scheduling, and smart city energy management, providing theoretical support and implementation tools for the active optimization of multi-source collaborative scheduling strategies; the active update determines that the strategy fails and actively updates when the comprehensive fitness decays to the threshold by setting the fitness threshold, and can make timely adjustments when the strategy gradually becomes unsuitable for the environment; the passive update immediately forces an update when the impact factor of price fluctuation or meteorological mutation exceeds the preset trigger value, and can quickly respond to the drastic changes in the external environment. The dual-insurance update mechanism ensures that the scheduling strategy can timely adapt to various changes inside and outside the system, improving the stability and reliability of the operation of the regional integrated energy system.

[0126] In some embodiments of the present invention, step S3 extracts the duration characteristics of the strategy persistence through the analysis of the strategy update time node, and the specific implementation is as follows:

[0127] Step S31: Collect multiple consecutive policy update time nodes. There are clear requirements for the collected time nodes, which should at least cover the current policy update time node, the next policy update time node, and the historical policy update time node that is closest to the current policy update time node in the time dimension. For example, if a policy update is in progress at the current moment, then the current policy update time node is the current moment; the next policy update time node is a future moment generated by the policy update node analysis model in step S2; and the historical policy update time node that is closest to the current policy update time node in the time dimension refers to the time point of the policy update that was closest to the current moment before this update. By collecting the above time nodes, a continuous time node sequence with a chronological order can be constructed, providing a data basis in the time dimension for subsequent analysis.

[0128] Step S32: Define the concept of policy duration, which represents the time span between two adjacent policy update time nodes. For example, if the previous policy update time node is t1 and the current policy update time node is t2, then the policy duration before this update is t2 - t1. Similarly, if the next policy update time node is t3, then the policy duration after this update is t3 - t2.

[0129] Step S33: Since multiple consecutive policy update time nodes are collected, multiple policy durations can be obtained. The duration data should at least include the policy duration before this update and the policy duration after this update. By obtaining multiple policy durations, the changes in the scheduling policy can be analyzed from the time dimension. For example, by comparing different policy durations, the stability of the policy in different time periods can be understood. If the policy duration is long in a certain period, it indicates that the scheduling policy is relatively stable during that period and can better adapt to the internal and external environment at that time. On the contrary, if the policy duration is short, it may mean that the current scheduling policy is difficult to adapt to environmental changes and needs further optimization. At the same time, the data of multiple policy durations also provide specific analysis objects for eliminating environmental external factors and evaluating stability in subsequent step S4, which helps to more comprehensively and deeply understand the performance and characteristics of the regional integrated energy system scheduling policy.

[0130] In some embodiments of the present invention, in order to strip the influence of external shocks such as electricity price, gas price, and meteorology and accurately evaluate the robustness of the scheduling policy under the action of internal equipment and load demand, the specific evaluation is as follows:

[0131] Step S41: Separate the external environmental noise in the policy duration and retain the pure policy duration dominated by the internal operating characteristics (equipment status, load demand) of the system. Establish an elimination mathematical model, and the input data includes:

[0132] The sequence of strategy duration is {t1, t2,..., t n}, where t i represents the duration of the i-th strategy update;

[0133] The sequence of external shock factors includes the price shock factor {ΔC price (1) , ΔC price (2) ,..., ΔC price (n)} and the meteorological shock factor {ΔE weather (1) , ΔE weather (2) ,..., ΔE weather (n)};

[0134] The elimination model uses multiple linear regression to establish a quantitative relationship between the strategy duration and the external shock factors:

[0135] ;

[0136] β0, β1, β2 represent the regression coefficients, reflecting the linear impact of external shocks on the duration;

[0137] represents the residual term, indicating the duration fluctuation that cannot be explained by external shocks in the i-th strategy update, that is, the pure strategy duration after eliminating external factors;

[0138] Calculate the duration after elimination:

[0139] ;

[0140] After eliminating the influence of external fluctuations such as electricity price and meteorology, the remaining duration reflects the effectiveness of the strategy itself.

[0141] Step S42: Based on the sequence of strategy durations {t1 adjusted , t2 adjusted ,..., t n adjusted} after eliminating external factors, calculate the stability index to quantify the stability of the strategy. The calculation formula is as follows:

[0142] CV = 1 - σ(t adjusted ) / μ(t adjusted );

[0143] where CV represents the stability index, σ(t adjusted ) represents the standard deviation of the duration after elimination; μ(t adjusted) represents the mean of the duration after elimination; the smaller the stability index, that is, the closer the stability index is to 1, the lower the volatility of the strategy duration and the higher the stability; if the duration after elimination is completely constant (σ = 0), the stability index is 1, indicating absolute stability.

[0144] In this embodiment, by eliminating and stripping external environmental factors such as electricity prices and meteorology, and then quantifying the stability of the strategy itself through the coefficient of variation, it provides accurate input for the model optimization in step S5; solves the lag problem of the fixed update cycle strategy, and ensures the efficiency and robustness of the scheduling strategy in the dynamic changes of the internal and external environments.

[0145] In some embodiments of the present invention, how to adjust and optimize the existing scheduling model according to the stability evaluation result of the strategy, and reapply the improved model to the system is as follows:

[0146] Step S51: Compare the stability index calculated in step S4 with a preset stability threshold; if the stability index is lower than the threshold, it indicates that the stability of the scheduling strategy generated by the current multi - collaborative scheduling model is insufficient and needs to be optimized; the greater the difference between the two, the higher the priority of optimization; conversely, if the stability meets the standard, no adjustment is temporarily required.

[0147] Step S52: When it is necessary to optimize the multi - collaborative scheduling model, adopt the following optimization strategies:

[0148] Optimize the supply - demand interaction game model, and the optimization goal is to improve the flexibility of equipment output and the matching degree with load demand; by updating the cost function in the game strategy and introducing a stability penalty term: new cost function = original cost function + λ·∣ΔS∣; where λ is the penalty coefficient, which is dynamically adjusted by the Lagrange multiplier method; ΔS represents the difference between the stability index and the preset stability threshold;

[0149] Optimize the user group evolutionary game model, and the optimization goal is to enhance the compliance and stability of users' response to instructions; for example, if it is required to reduce electricity consumption by 20%, but actually only 15% is reduced, the system will reduce the priority weight of such users and reduce the dependence on them; by introducing a reputation mechanism, users with long - term cooperation are given higher priority. For example, if a factory has responded well in the past, subsequent instructions will be preferentially allocated to it, thereby improving the overall execution rate. The reputation value calculation formula is:

[0150] Reputation value = α·historical response rate+(1 - α)·recent response rate;

[0151] Among them, α is the attenuation factor;

[0152] Optimize the benefit evaluation model, with the optimization goal of balancing economy and stability; find the optimal solution through optimization algorithms, such as multi-objective genetic algorithms, to achieve a balance between reducing energy costs (such as purchasing electricity during low-price periods) and enhancing stability; for example, the algorithm may recommend a slight increase in the gas procurement cost, but it can reduce the strategy duration fluctuation by 30%, thus significantly enhancing stability.

[0153] Step S53: The optimized model will not be put into use immediately. Instead, it will first be simulated and run through historical data to ensure that key indicators, such as stability and cost, are improved; after passing the verification, the system will adopt a progressive update strategy, which is as follows:

[0154] Run the new and old models simultaneously and compare the real-time effects; for example, the new model is responsible for area A and the old model is responsible for area B, and observe whether the stability of area A is improved;

[0155] After confirming that the new model is effective, seamlessly replace the old model through background technology to avoid service interruption;

[0156] If the stability unexpectedly decreases after the update, such as due to data anomalies, the system will automatically switch back to the old version to ensure operation safety.

[0157] In this embodiment, the adaptability and reliability of the regional integrated energy system are significantly improved through a dynamic optimization closed-loop mechanism; based on the stability difference, trigger directional adjustments to avoid resource waste and ensure that the optimization focuses on key issues; the supply-demand model introduces a stability penalty term to improve the equipment regulation ability and load matching efficiency; the user model strengthens the weight of high-response users through a reputation mechanism to improve the execution rate; the benefit model adopts a multi-objective optimization algorithm to balance economy and stability and avoid neglecting a single objective; through A / B test verification, hot update, and automatic rollback mechanisms, ensure zero downtime for policy updates and reduce implementation risks; combine real-time evaluation and dynamic optimization to form a closed loop of perception, analysis, decision-making, and verification, continuously adapt to complex environmental changes, and achieve long-term stable and efficient operation of the system.

[0158] In some solutions, multiple embodiments of the present application can be combined and the combined solutions can be implemented. Optionally, some operations in the processes of the method embodiments are optionally combined, and / or the order of some operations is optionally changed. And, the execution order between the steps of each process is only exemplary and does not constitute a limitation on the execution order between the steps. There can also be other execution orders between the steps. It is not intended to indicate that the execution order is the only order in which these operations can be performed. Those of ordinary skill in the art will think of various ways to reorder the operations described herein. Additionally, it should be noted that the process details involved in a certain embodiment herein are equally applicable to other embodiments in a similar manner, or different embodiments can be combined and used.

[0159] In addition, some steps in the method embodiments can be equivalently replaced with other possible steps. Or, some steps in the method embodiments can be optional and can be deleted in some usage scenarios. Or, other possible steps can be added to the method embodiments. Moreover, the method embodiments can be implemented independently or in combination with each other.

[0160] The various variations and specific embodiments of the multi-source supply collaborative scheduling method for the regional integrated energy system in the foregoing embodiments are equally applicable to the multi-source supply collaborative scheduling system for the regional integrated energy system in this embodiment. Through the foregoing detailed description of the multi-source supply collaborative scheduling method for the regional integrated energy system, those skilled in the art can clearly know the implementation method of the multi-source supply collaborative scheduling system for the regional integrated energy system in this embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail herein.

[0161] The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only to illustrate the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi - supply collaborative scheduling system for regional integrated energy systems, characterized in that, Including: A data acquisition module, configured to acquire in real time a device operation status data set and an energy consumption load demand data set in a regional integrated energy system; A multi - element collaborative strategy generation module, which uses a built - in multi - element collaborative scheduling model to operate on the device operation status data set and the energy consumption load demand data set to obtain a multi - element collaborative scheduling strategy; A strategy update module, configured to acquire external environment information in real time, and input it and the multi - element collaborative scheduling strategy into a preset strategy update node analysis model to generate the next strategy update time node; A duration feature extraction module, configured to collect multiple consecutive strategy update time nodes, and perform duration feature extraction to obtain multiple strategy duration lengths; A stability evaluation module, configured to eliminate external environmental factors for each of the strategy duration lengths, and perform stability evaluation on the multiple strategy duration lengths after elimination to obtain a stability index; A model optimization and deployment module, configured to optimize the multi - element collaborative scheduling model according to the difference between the stability index and a preset stability threshold, and update and deploy the optimized model to the regional integrated energy system.

2. The multi-source supply collaborative scheduling system for regional integrated energy system according to claim 1, wherein The duration feature extraction module is further configured as: The multiple consecutive strategy update time nodes at least include the current strategy update time node, the next strategy update time node, and the historical strategy update time node that is closest to the current strategy update time node in the time dimension.

3. A multi-source supply collaborative scheduling method for regional integrated energy systems, characterized in that, Including: Acquire in real time a device operation status data set and an energy consumption load demand data set in a regional integrated energy system, and input both into a multi - element collaborative scheduling model to obtain a multi - element collaborative scheduling strategy; Acquire external environment information in real time, and input it and the multi - element collaborative scheduling strategy into a strategy update node analysis model to generate the next strategy update time node; Collect multiple consecutive strategy update time nodes, and perform duration feature extraction to obtain multiple strategy duration lengths; Eliminate external environmental factors for each of the strategy duration lengths, and perform stability evaluation on the multiple strategy duration lengths after elimination to obtain a stability index; Optimize the multi - element collaborative scheduling model based on the difference between the stability index and a preset stability threshold, and update and deploy the optimized multi - element collaborative scheduling model to the regional integrated energy system.

4. The multi-source supply collaborative scheduling method for regional integrated energy systems according to claim 3, wherein, The device operation status data set includes operation status information of various energy supply devices, and the energy consumption load demand data set includes real - time energy consumption demand information of different types of energy - consuming load groups.

5. The multi-source supply collaborative scheduling method for regional integrated energy system according to claim 4, characterized in that The multi - element collaborative scheduling model adopts a composite model, including a supply - demand interaction game model, a user group evolution game model, and a benefit evaluation model.

6. The multi - supply collaborative scheduling method for regional integrated energy systems according to claim 5, wherein, The external environment information includes grid electricity price, natural gas price, and meteorological data.

7. The multi - supply collaborative scheduling method for regional integrated energy system according to claim 6, characterized in that, The multiple consecutive strategy update time nodes at least include the current strategy update time node, the next strategy update time node, and the historical strategy update time node that is closest to the current strategy update time node in the time dimension.

8. The multi-source supply collaborative scheduling method for regional integrated energy systems according to claim 7, characterized in that The strategy duration length represents the time span between two adjacent strategy update time nodes; The durations of the multiple strategies at least include the duration of the strategy before the current update and the duration of the strategy after the current update.

9. The multi-source supply collaborative scheduling method for a regional integrated energy system according to any one of claims 3-8, characterized in that The active update trigger condition of the strategy update node analysis model is: when the comprehensive fitness of the current multi - collaborative scheduling strategy decays to the set fitness threshold, it is determined that the current multi - collaborative scheduling strategy fails and needs to be actively updated; The formula for calculating the comprehensive fitness is: ； Among them, F(t) represents the comprehensive fitness; t represents the duration of the current multi - collaborative scheduling strategy; F0 represents the normalized initial fitness, indicating that the initial state of the strategy is optimal; -ΔC price ·t represents that the price shock decays linearly with time. The greater the shock, the faster the comprehensive fitness decreases; -ΔE weather ·t 2 represents that the meteorological shock decays with the square of time, reflecting the non - linear impact of emergencies; A s +A d represents the weighted sum of the adaptabilities of the supply side and the demand side. The larger the value, the more robust the strategy is; By inversely solving the duration of the current multi - collaborative scheduling strategy through the set fitness threshold and combining the update time node of the current strategy, the next update time node is calculated.

10. The multi-source supply collaborative scheduling method for regional integrated energy system according to claim 9, characterized in that, The said ΔC price represents the impact factor of price fluctuation, and its calculation formula is: ； Among them, P e (t) represents the electricity price at time t; P e (t - Δt) represents the electricity price at the previous monitoring time point (t - Δt); P g (t) represents the natural gas price at the current time t, with the unit of yuan / m³; P g (t - Δt) represents the natural gas price at the previous monitoring time point (t - Δt); d1 max represents the upper limit of the electricity price peak in the floating range of time-of-use electricity price on the demand side; d1 min represents the lower limit of the electricity price valley in the floating range of time-of-use electricity price on the demand side; P g base represents the historical benchmark value of the natural gas price, which is used to measure the degree to which the current gas price deviates from the historical average.

Images