Distributed resource peak shaving market electric quantity declaration method and system, electronic equipment and medium

Abstract

The present application relates to the technical field of power distributed resource regulation, and particularly relates to a distributed resource peak regulation market power reporting method and system, electronic equipment and medium. The method first acquires real-time operation data of load side distributed resources and peak regulation market policy data; constructs a double-constraint power generation power model according to the real-time operation data and the peak regulation market policy data; constructs a distributed resource aggregation model based on the double-constraint power generation power model; extracts non-steady variables from the resource aggregation state and the power reporting reference value; obtains an optimal power reporting strategy by using interval optimization logic according to the non-steady variables; and finally generates a power reporting scheme for participating in the peak regulation market. Through the double-constraint power generation power model and the interval optimization logic, the problem of insufficient optimization of peak regulation market benefits caused by ignoring the adjustable load characteristics and wind and light curtailment penalties in the prior art is solved, and the adaptability and reliability of the power reporting strategy in a complex market environment are improved.

Description

Technical Field

[0001] This invention relates to the field of power distributed resource regulation technology, and in particular to a method, system, electronic device and medium for reporting peak-shaving market electricity volume for distributed resources. Background Technology

[0002] With the increasing penetration of new energy sources in the power system, the demand for peak shaving in the power grid is growing. As an important component of the ancillary services market, the peak shaving market places high demands on the regulating capacity of participating entities (typically 5 MW or 10 MW, varying by province). Distributed resources, especially adjustable load resources on the user side, are characterized by local production, scattered distribution, and small single-unit capacity. Individual distributed energy sources often cannot meet the entry threshold of the peak shaving market. Therefore, aggregation technology is needed to integrate dispersed resources to participate in the market and obtain revenue. In existing technologies, resource aggregators typically employ strategies such as regional aggregation, energy type aggregation, or budget-driven aggregation to package distributed resources as reference values ​​for quantity and price declarations, and optimize the output scheduling of each resource based on overall revenue or grid security constraints.

[0003] However, existing aggregation and optimization methods have significant limitations. For example, prior art document 1 (CN111762051A) discloses a virtual power plant optimization scheduling method, which focuses on allocating the output of photovoltaic and energy storage resources through a cost model. However, this method relies heavily on static aggregation strategies and fails to fully consider the dynamic characteristics of adjustable load resources (such as output fluctuations caused by weather) and the real-time impact of wind and solar curtailment penalty costs on revenue. Specifically, adjustable load resources depend on demand-side response and are easily affected by environmental factors (such as temperature and sunlight). If such resources are blindly aggregated without incorporating their uncertainty modeling, it may lead to a deviation between the declared power volume and the actual peak-shaving task, thereby triggering deviation assessment fees. At the same time, the wind and solar curtailment penalty cost, as a key factor affecting revenue, is often simplified to a fixed value in existing technologies and is not coupled with real-time market price signals, making it difficult to dynamically optimize revenue.

[0004] Therefore, existing technologies struggle to maximize the benefits of distributed resource aggregation in complex market environments. They lack multi-dimensional dynamic consideration of the characteristics of adjustable load resources and the cost of wind and solar curtailment penalties, resulting in insufficient adaptability and economic efficiency of aggregation solutions in fluctuating scenarios, and failing to guarantee the optimization of peak-shaving market benefits. Summary of the Invention

[0005] To address the aforementioned shortcomings or deficiencies, this invention provides a method, system, electronic device, and medium for reporting peak-shaving electricity in the distributed resource market. This solves the problem of insufficient optimization of peak-shaving market revenue caused by neglecting adjustable load characteristics and wind and solar curtailment penalties in existing technologies.

[0006] This invention provides a method for reporting peak-shaving market electricity demand using distributed resources, comprising:

[0007] Obtain operational data of distributed resources on the load side to obtain information on peak shaving market policies.

[0008] A dual-constraint power generation model is constructed based on operational data and peak-shaving market policy information.

[0009] Based on the dual-constraint power generation model, a distributed resource aggregation model is constructed by integrating resource characteristic parameters and wind and solar curtailment penalty parameters. This model is used to generate resource aggregation status and power application reference values.

[0010] Unsteady variables are extracted from resource aggregation status and electricity declaration reference values, and then transformed into interval numbers.

[0011] Based on the transformed unsteady variables, the optimal electricity declaration strategy is obtained by using interval optimization logic to generate an electricity declaration plan for participating in the peak shaving market.

[0012] According to a second aspect, this invention provides a distributed resource peak-shaving market electricity reporting system, comprising:

[0013] The operation and market data acquisition module is used to acquire operation data of distributed resources on the load side to obtain information on peak shaving market policies.

[0014] The power generation model building module is used to build a dual-constraint power generation model based on operating data and peak-shaving market policy information.

[0015] The resource aggregation model construction module is used to build a distributed resource aggregation model based on the dual-constraint power generation model, integrating resource characteristic parameters and wind and solar curtailment penalty parameters. The distributed resource aggregation model is used to generate resource aggregation status and power declaration reference values.

[0016] The unsteady variable extraction module is used to extract unsteady variables from resource aggregation status and electricity declaration reference values, and convert the unsteady variables into interval numbers.

[0017] The electricity declaration strategy generation module is used to obtain the optimal electricity declaration strategy based on the transformed unsteady variables using interval optimization logic, so as to generate an electricity declaration scheme for participating in the peak shaving market.

[0018] According to a third aspect, the present invention provides an electronic device comprising:

[0019] At least one processor; and

[0020] The memory that is communicatively connected to the at least one processor;

[0021] The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the peak-shaving market electricity reporting method for any distributed resource in the embodiments of the present invention.

[0022] According to another aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause a computer to execute a peak-shaving market electricity reporting method for any distributed resource in the embodiments of the present invention.

[0023] This invention provides a method for reporting electricity demand in the peak-shaving market for distributed resources. This method is executed through an integrated data acquisition and optimization calculation module. The method includes: acquiring real-time operational data of distributed resources on the load side and peak-shaving market policy data, and performing consistency verification and standardization preprocessing on the multi-source data. Next, based on the preprocessed real-time operational data and peak-shaving market policy data, a dual-constraint power generation model is constructed. This model characterizes the power generation calculation rules that simultaneously satisfy resource output capacity boundaries and market compliance boundaries. Then, based on the dual-constraint power generation model, resource characteristic parameters and wind / solar curtailment penalty parameters are integrated to construct a distributed resource aggregation model. Further, non-stationary variables are extracted from the resource aggregation status and electricity reporting reference values, and these non-stationary variables are transformed into interval numbers to encapsulate volatility. Then, based on the transformed non-stationary variables, interval optimization logic is used to solve for the optimal electricity reporting strategy. Finally, based on the optimal electricity reporting strategy, a reporting scheme conforming to peak-shaving market regulations is generated.

[0024] Throughout the process, this invention addresses the problem described in the background art of "failing to consider the characteristics of adjustable load resources and the penalties for wind and solar curtailment in relation to revenue." By constructing a dual-constraint power generation model and a distributed resource aggregation model, it achieves a multi-dimensional integrated analysis of the dynamic characteristics of adjustable load resources and the costs of wind and solar curtailment penalties, overcoming the limitations of traditional methods that only focus on the overall resource capacity. Furthermore, through interval number transformation and interval optimization logic, it effectively handles uncertainties such as market electricity price fluctuations and resource output deviations, ensuring the robustness and economic efficiency of the electricity reporting strategy. Therefore, the technical solution of this invention solves the problem of insufficient optimization of peak-shaving market revenue caused by neglecting the characteristics of adjustable loads and the penalties for wind and solar curtailment in existing technologies, improving the adaptability and reliability of the electricity reporting strategy in complex market environments. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of a system architecture for distributed resource aggregation participating in the peak-shaving market according to an embodiment of the present invention;

[0026] Figure 2This is a flowchart of a method for reporting peak-shaving market electricity consumption using distributed resources according to an embodiment of the present invention;

[0027] Figure 3 This is an optimized flowchart of a distributed resource aggregation model according to an embodiment of the present invention;

[0028] Figure 4 This is a structural block diagram of a distributed resource peak-shaving market electricity reporting system according to an embodiment of the present invention;

[0029] Figure 5 This is a block diagram of an electronic device used to implement embodiments of the present invention. Detailed Implementation

[0030] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0031] This invention provides a method for reporting electricity demand in the peak-shaving market for distributed resources, based on a first aspect. This method can be applied to a system for reporting electricity demand in the peak-shaving market for distributed resources (hereinafter referred to as the "system"). The system runs on a distributed resource aggregation platform through local deployment or containerization to automate the entire process from data collection and modeling optimization to the generation of electricity demand reporting strategies.

[0032] Specifically, the physical devices of distributed resources include, but are not limited to, smart meters deployed on the load side, environmental sensor networks, terminal data acquisition devices, and edge computing gateways. These physical devices need to have real-time data acquisition, protocol conversion, and edge computing capabilities to support millisecond-level data acquisition frequencies and minute-level model update cycles, ensuring that the system can respond promptly to peak-shaving market fluctuations.

[0033] In such Figure 1 The system architecture diagram shown illustrates distributed resource aggregation participating in the peak-shaving market. Physical devices are connected to the aggregation platform via IoT gateways. The platform incorporates three core modules: a dual-constraint power generation model, a distributed resource aggregation model, and an interval optimization algorithm. This architecture supports both independent single-machine deployment and elastic scaling across multiple nodes through containerization technology, meeting the business needs of distributed resource aggregators of different sizes.

[0034] The system securely interacts with the peak-shaving market center through the electricity market data interface, ultimately outputting a compliant electricity declaration scheme. Generally, the physical equipment needs to support industrial protocols such as Modbus and IEC 104, as well as communication protocols such as HTTPS (Hypertext Transfer Protocol Secure) or MQTT (Message Queuing Telemetry Transport), to ensure reliable end-to-end transmission and processing from data collection to declaration decision-making.

[0035] like Figure 2 As shown, the method may include:

[0036] Step S110: Obtain operational data of distributed resources on the load side to obtain information on peak shaving market policies.

[0037] Distributed resources on the load side refer to distributed energy devices located on the electricity user side, such as adjustable loads, distributed photovoltaics, and energy storage devices. These resources are characterized by their scattered distribution and small individual unit capacity. Operational data includes real-time measured physical quantities, such as the regulating capacity of adjustable loads (unit: kilowatts), environmental monitoring data (such as temperature and humidity, unit: degrees Celsius), and equipment operating status data (such as switch status). Peak-shaving market policy information includes the rules and parameters of the peak-shaving ancillary services market, such as market access power (unit: kilowatts), deviation assessment coefficient (dimensionless), and real-time electricity price (unit: yuan per kilowatt-hour).

[0038] Specifically, the system can collect regulation capacity data through smart meters deployed on the load side, with a sampling frequency of once per minute; then, it acquires environmental monitoring data through an environmental sensor network, using the MQTT transmission protocol; finally, it obtains policy information through an external electricity market data interface, using the HTTPS communication protocol to ensure data security. The steps of collecting various types of information can be executed in parallel, sequentially, or in different orders.

[0039] For example, the system collects real-time power data of adjustable loads in an industrial park, with a maximum adjustment capacity of 500 kW and a minimum adjustment capacity of 100 kW; at the same time, it obtains the access power threshold of the peak-shaving market in the region as 1000 kW, the deviation assessment coefficient as 0.05, and the real-time electricity price as 0.8 yuan per kilowatt-hour.

[0040] Step S120: Construct a dual-constraint power generation model based on operating data and peak-shaving market policy information.

[0041] Among them, the dual-constraint power generation model is a mathematical model used to calculate the range of power generation that distributed resources can declare in the peak-shaving market, while satisfying both the resource output capacity boundary (determined by physical operating characteristics) and the market compliance boundary (determined by policy rules).

[0042] Specifically, the system first extracts the maximum and minimum regulation capacities from operational data to form a set of physical operational characteristic parameters. Next, it extracts market access power and deviation assessment coefficients from policy information to form a set of market policy constraint parameters. Then, based on these two parameter sets, it calculates the declared power range using a linear programming algorithm and performs dynamic threshold verification, with the verification boundary adjusted in real time according to environmental data.

[0043] For example, the system takes the maximum adjustable capacity of the adjustable load of 500 kW and the market access power of 1000 kW as input, calculates the declared power range of 100 kW to 500 kW, and verifies whether the range meets the deviation assessment requirements (such as power deviation not exceeding 5%).

[0044] In some embodiments, the adjustment magnitude of adjustable load resources needs to be comprehensively considered based on the supply and demand situation in the peak-shaving market and the agreed price between the aggregator and the load-side adjustable resources. The system can model the dual-constraint power generation model using the following formula:

[0045] ;

[0046] in, The actual load reduction (in kilowatts) provided in the j-th step at time t. and These represent the maximum and minimum regulating capacity (in kilowatts) that can be provided at this moment and in this step, respectively. The system can obtain the maximum and minimum regulating capacity from the load side in real time through smart meters and terminal acquisition devices, with a sampling frequency of milliseconds, supporting flexible control of the equipment. This represents the load reduction (in kilowatts) already provided by the load side at time t, used to accumulate historical regulation effects.

[0047] Furthermore, adjustable load resources must adhere to specific operational constraints when participating in the peak-shaving market. The following formula describes the linear relationship between adjustable load and total load:

[0048] ;

[0049] in, It can represent the adjustable load or demand response at time interval t, which belongs to the load power that can be directly controlled or stimulated to be adjusted by the energy management system. It can represent the total load or baseline load at time segment t, and is used to characterize the total power demand of the system at that moment. It is a proportionality coefficient, and its value range is generally 0≤η≤1, representing the proportion of adjustable load to total load.

[0050] For example, the proportion of adjustable load to total load (denoted as η) is a key constraint parameter. This proportion is dynamically adjusted based on the date type. For instance, the system sets a higher value (e.g., 0.7, or 70%) on weekdays to match the high load demand of commercial and industrial electricity use; it is appropriately reduced on ordinary rest days (e.g., 0.5, or 50%) to reflect changes in residential electricity consumption patterns; and it is further adjusted on holidays (e.g., 0.4, or 40%) to adapt to holiday load characteristics. This dynamic adjustment mechanism can be achieved through historical load data analysis and machine learning algorithms (such as time series forecasting). The system automatically identifies the date type and updates the proportion value to ensure that the constraint conditions are consistent with the actual operating environment in real time. In addition, the proportion adjustment must also take into account grid safety rules, such as avoiding overload or voltage exceeding limits, and is verified through the Energy Management System (EMS).

[0051] Step S130: Based on the dual-constraint power generation model, integrate resource characteristic parameters and wind and solar curtailment penalty parameters to construct a distributed resource aggregation model.

[0052] Among them, the distributed resource aggregation model is used to generate resource aggregation status and power application reference value. The resource aggregation status is used to represent whether the distributed resource is selected to participate in the aggregation (such as binary identifier, 0 means not selected, 1 means selected). The power application reference value represents the recommended power application amount (unit: kilowatt-hour).

[0053] In this embodiment, resource characteristic parameters include resource type weighting (e.g., adjustable load weighting is 0.6, photovoltaic weighting is 0.4) and resource type identifier (e.g., classification code); the wind and solar curtailment penalty parameter includes the unit curtailment penalty price (unit: yuan per kilowatt-hour). This resource type weighting can be dynamically initialized based on historical output data or expert experience.

[0054] Specifically, the system can construct an aggregate objective function with the goal of minimizing the total aggregate cost. The total cost includes resource operation costs, wind and solar curtailment penalty costs, and market transaction costs. The constraints include upper limit constraints on power deviation (e.g., the deviation between actual output and declared value does not exceed 10 kilowatts) and adjustable load ratio constraints (e.g., the adjustable load ratio does not exceed 30% of the total load).

[0055] For example, the system aggregates a set of distributed resources, including adjustable load (weight 0.6) and photovoltaic (weight 0.4), and the objective function minimizes the total cost (e.g., operating cost of 1000 yuan and penalty cost of 200 yuan). The system outputs the resource aggregation status (adjustable load is selected, photovoltaic is not selected) and the reference value for electricity declaration (e.g., declaration of 50 kWh).

[0056] In some embodiments, when constructing a distributed resource aggregation model, the design of the objective function and constraints aims to optimize the resource aggregation strategy to minimize the total cost, while ensuring market compliance and operational feasibility. Specifically, the aggregation objective function is defined as minimizing the total cost of participating in the peak-shaving market, and its expression is as follows:

[0057] ;

[0058] in, ~ These represent the aggregation state of the i-th type of distributed resource (with values ​​of 0 or 1, where 1 indicates participation in aggregation). ~ These represent the weighting of resource types, which are derived from historical data; i=1, 2, 3 correspond to energy types, where i=1 and 2 can represent new energy sources such as photovoltaic and wind power, and i=3 represents adjustable load resources. , , These are the economic costs of the corresponding resources. This indicates the penalty cost for abandoning wind and solar power. and These represent the purchase price and the amount of electricity purchased when participating in peak-shaving transactions from the market.

[0059] Specifically, in the optimization model of the electricity application method in the distributed resource peak-shaving market, Represents the cost of traditional power generation. In a resource aggregation state, These are the weighting coefficients. (for power generation costs). Indicates the fixed cost of distributed generation. and For the aggregation state and weight of the corresponding resources, For the cost of electricity generation; For demand response costs, In adjustable load aggregation state, These are the weighting coefficients; ,in It is the demand response compensation price for time period t. For response quantity; The cost of curtailment penalties for wind and solar power (determined by the solar power curtailment penalty price) Multiply by the difference between available photovoltaic power and actual output. And the penalty price for wind curtailment Multiply by the difference in wind power ( (together calculated); last item This represents the cost of purchasing electricity from the main grid. For electricity purchase price, This function calculates the electricity purchase amount. By weighting and integrating physical operating costs, market penalty costs, and external electricity purchase costs, it provides a quantitative economic target for distributed resource aggregation to participate in the peak-shaving market.

[0060] In the above embodiments, the aggregation constraints may include the following constraints (1) and (2):

[0061] Power deviation constraint (1): ,in The power limit for new energy sources participating in the peak-shaving market is generally not to exceed the total peak-shaving task, ensuring that the deviation between the actual output and the declared value is within the allowable range.

[0062] Adjustable load ratio constraint (2): Where η is the proportion of adjustable load to total load, dynamically adjusted according to date types such as weekdays, rest days, and holidays to adapt to changes in load characteristics. By integrating the above objective function and constraints, the system can quantify the impact of resource characteristics and penalty costs, generate optimal resource aggregation state and electricity reporting reference values, thereby reducing market penalty risk and improving revenue stability.

[0063] Step S140: Extract non-steady variables from resource aggregation status and electricity declaration reference values, and convert the non-steady variables into interval numbers.

[0064] Among them, unsteady variables refer to variables with uncertainty in the peak-shaving market, such as the adjustable load output forecast deviation (unit: kilowatt) and market electricity price fluctuation (unit: yuan per kilowatt-hour); the interval number form uses intervals to represent the possible range of variables, such as [lower limit, upper limit].

[0065] Specifically, the system uses Latin hypercube sampling (LHS, a statistical sampling method) to generate multiple sets of sample datasets, covering the possible value range of non-stationary variables; then it calculates the upper and lower bounds of the interval numbers and adjusts the interval boundaries through a dynamic correction strategy (such as rolling updates based on historical data), with a correction cycle of 15 minutes.

[0066] For example, for the adjustable load output prediction deviation, 100 sets of data are sampled and generated, with a calculation range of [-10, 10] kW; for market electricity price fluctuations, the range is [0.5, 1.0] yuan per kilowatt-hour; after dynamic correction, the range is updated to... kilowatts and Yuan per kilowatt-hour.

[0067] For example, the sampling process is implemented using a Python library (such as SciPy), and the dataset size is 1000 records.

[0068] For example, the dynamic correction strategy updates the interval using a sliding window algorithm by comparing the deviation between the actual value and the predicted value (e.g., triggering an adjustment when the deviation exceeds 5%).

[0069] Step S150: Based on the transformed unsteady variables, use interval optimization logic to solve for the optimal electricity declaration strategy, so as to generate an electricity declaration plan for participating in the peak shaving market.

[0070] Among them, interval optimization logic is a mathematical optimization method used to process the objective function of interval numbers and find the optimal solution by balancing the expected value and stability; the optimal electricity declaration strategy includes the declared electricity value (unit: kilowatt-hour) and the declaration time series (such as a list of time points); the electricity declaration scheme is a standardized output that conforms to market norms, such as a JSON (a lightweight data exchange format) file.

[0071] For example, the system constructs an interval number objective function with an objective tradeoff coefficient of 0.7 (indicating a greater emphasis on stability), and uses a reinforcement learning algorithm (such as Q-learning) for iterative optimization. The convergence condition is that the cost change is less than 1% for 10 consecutive iterations. The output is 100 kWh of electricity to be reported, and the time series is reported once per hour (e.g., 50 kWh reported at 10:00 and 50 kWh reported at 11:00).

[0072] In some embodiments, when solving for the optimal scheduling scheme, the system addresses the volatility and intermittency of adjustable resources due to weather conditions by employing an interval optimization method to transform uncertainty into a deterministic decision problem. Specifically, interval numbers are used to represent the possible range of output from adjustable resources on the load side, encapsulating unsteady variables such as forecast bias and market fluctuations. The interval optimization model can be formally represented as follows:

[0073] ;

[0074] in, Let U be the matrix of decision variables (e.g., reported electricity volume, resource dispatch instructions), and U be the matrix of uncertain variables (e.g., adjustable load output deviation, market electricity price fluctuations); I and q represent the number of uncertain constraints and the number of uncertain variables, respectively, corresponding to the adjustable output on the load side and the number of equipment. Specifically, This means that the i-th constraint function (such as power balance or equipment output limit) must be less than or equal to an interval constant. (Its upper and lower limits are) and This reflects the potential ambiguity or fluctuation range of the constraint boundaries in actual operation; uncertain variables U (such as renewable energy output and market electricity prices) and each of their components. They are all confined to their respective ranges. = and Within this framework, uncertainty is quantified; while the decision variable X (such as resource aggregation status and declared electricity volume) needs to be selected from the n-dimensional feasible region. This model transforms uncertainty into deterministic decisions through an interval ranking relationship comparison method to cope with real-time electricity price fluctuations.

[0075] For example, the constraints include power balance constraints expressed by the following formula:

[0076] ;

[0077] in, and These are the power generation from wind power and solar power, respectively. To regulate the power supply on the load side, Purchase electricity for the power grid Adjustable power supply for distributed resources. The model allows users to submit electricity demand for peak shaving. It transforms uncertain constraints into deterministic forms using interval mathematics (such as interval ranking relationships and probability comparisons), for example, by introducing upper and lower bounds and dynamically correcting them to ensure the robustness of the optimization results. Ultimately, the system generates the optimal electricity submission strategy, improving the stability and adaptability of peak shaving market revenue. Specifically, interval mathematics is a mathematical method for handling uncertainty, used to represent variables as interval numbers. The possibility degree method is a commonly used technique in interval mathematics and fuzzy optimization, used to compare the priority of interval numbers or fuzzy numbers, especially when dealing with uncertain optimization problems. This method evaluates the relative merits of different optimization schemes by calculating the degree of overlap between intervals, possibility values, or other metrics. A standard and widely cited textbook on the possibility degree method is "Fuzzy Sets and Systems: Theory and Applications," a classic work in the field of fuzzy set theory and system applications. Chapter 6, "Possibility Theory and Applications" (approximately pages 120-150), elaborates on the basic principles of the possibility degree method and its implementation in uncertain optimization. Specifically, the book systematically introduces interval mathematics and fuzzy optimization techniques, including how to handle the interval number sorting problem through probability comparison. This provides theoretical support for the transformation of unsteady variables and robust decision-making in the interval optimization logic of this invention, and enhances the adaptability and reliability of the electricity reporting strategy in the uncertain environment of the peak-shaving market.

[0078] For example, when applying interval mathematics theory to transform the uncertainty in distributed resource optimization scheduling into a deterministic model, the system can perform two key steps to ensure robustness. First, in the uncertain constraint transformation step, interval numbers A and B (satisfying A≤B) are introduced, transforming the interval constraints containing uncertain parameters into the following deterministic form:

[0079] ;

[0080] ;

[0081] For each uncertainty constraint X is the decision variable matrix, U is the uncertain variable matrix, and the system pre-sets the tradeoff coefficients. (Values ​​range from 0 to 1, typically initialized based on historical data or expert experience), and calculate the range of values ​​for the constraint at X. ,in and These are the lower and upper bounds of the interval, respectively.

[0082] Secondly, in the transformation step of the uncertain objective function, when X is determined, the uncertain vector U causes the range of values ​​of the objective function f(X,U) to be encapsulated in the interval form shown below [ , ]express:

[0083] ;

[0084] ;

[0085] in, and These represent the upper and lower limits of the interval, respectively, and are solved using interval sequence relation methods (such as possibility comparison or interval order relation). Furthermore, the interval is decomposed into central values. and radius These are analogous to the expected value and standard deviation in stochastic programming, used to quantify the average performance and volatility of the objective function. Ultimately, the system is adjusted through iterative optimization (such as reinforcement learning algorithms). By defining the interval boundaries, a stable and optimally profitable electricity reporting strategy is generated, improving the adaptability of participation in the peak-shaving market.

[0086] In some embodiments, the optimization process of the distributed resource aggregation model is as follows: Figure 3 As shown, its core principle is to transform uncertainty into schedulable decisions through iterative learning and parameter correction, thereby ensuring policy robustness. The specific steps are as follows:

[0087] Step (1) Generate sample dataset:

[0088] The Latin hypercube sampling method is used to generate adjustable load data samples to ensure that the data uniformly covers the output range and avoids interference from extreme values. Historical load data is split according to dimensions such as peak-shaving market, power purchase behavior, and wind and solar curtailment. A sample set is constructed by combining real-time electricity price and preset cost as the model input.

[0089] Step (2) Correct the model weights:

[0090] The resource type weights λ_i are iteratively calculated based on historical data, the cost and benefit of the generation strategy are observed, and the weight parameters in the objective function are dynamically adjusted.

[0091] Step (3) Parameter correction and loop mechanism:

[0092] Compare the actual running results with the model output deviation, and use robust optimization to adjust the uncertainty set boundary. If the global optimal solution cannot be found (the "no" branch of the decision box), the system automatically enters the parameter correction stage, redefines the interval range, updates the corrected parameters to the objective function, and if the termination condition is not met, returns to step (1) to regenerate samples and form a closed-loop optimization.

[0093] This process uses an iterative structure of "generation-correction-adjustment-update" to enable the model to continuously adapt to market fluctuations and ultimately output an aggregation strategy that balances economy and robustness.

[0094] In the above embodiments, the system can also transform the uncertain objective function into a deterministic form using interval optimization theory, as shown in the following formula:

[0095] ;

[0096] Where X is a decision variable vector, representing optimizable variables such as resource aggregation status and reported electricity volume; It is the comprehensive objective function to be minimized; It usually refers to economic objectives (such as total system operating costs). This represents the stability or risk control objective (such as power fluctuation penalty or robustness cost); the core parameter φ is a tradeoff coefficient (0≤φ≤1), used to adjust the preference between economy and stability—when φ approaches 0, the optimization leans towards economy, and when it approaches 1, it focuses on stability. This formula transforms the multi-objective optimization problem into a single-objective problem through linear weighting, achieving a balanced optimization of different performance indicators.

[0097] Specifically, the deterministic form described above also includes the following constraints:

[0098] ;

[0099] The constraints are divided into two categories: the first category is probabilistic constraints, which represent the uncertain functions corresponding to the decision variable X. The value does not exceed the given threshold The probability must be no less than the preset confidence level. The first constraint applies to all uncertain scenarios in the model and can be used in actual scheduling to describe the reliability of adjustable load resource output in meeting peak-shaving requirements; the second line is a variable range constraint, specifying that the decision variable X belongs to the n-dimensional feasible solution space. Target tradeoff coefficient φ and confidence level The values ​​of are all within the interval [0,1] to ensure the mathematical rationality and practical solvability of the optimization problem. These constraints collectively define the feasible region of the model, ensuring that the generated scheduling scheme still meets the operational reliability requirements under uncertain environments.

[0100] Therefore, according to the above implementation method, firstly, real-time operational data of distributed resources on the load side and peak-shaving market policy data are acquired, and consistency verification and standardization preprocessing are performed on the multi-source data. Next, based on the preprocessed real-time operational data and peak-shaving market policy data, a dual-constraint power generation model is constructed. This model characterizes the power generation calculation rules that simultaneously satisfy resource output capacity boundaries and market compliance boundaries. Then, based on the dual-constraint power generation model, resource characteristic parameters and wind / solar curtailment penalty parameters are integrated to construct a distributed resource aggregation model. Further, non-stationary variables are extracted from resource aggregation status and electricity declaration reference values, and these non-stationary variables are transformed into interval numbers to encapsulate volatility. Then, based on the transformed non-stationary variables, interval optimization logic is used to solve for the optimal electricity declaration strategy. Finally, based on the optimal electricity declaration strategy, a declaration scheme conforming to peak-shaving market regulations is generated.

[0101] Throughout the process, this embodiment addresses the issue mentioned in the background art of "failing to consider the characteristics of adjustable load resources and the penalties for wind and solar curtailment in relation to revenue." By constructing a dual-constraint power generation model and a distributed resource aggregation model, it achieves a multi-dimensional integrated analysis of the dynamic characteristics of adjustable load resources and the costs of wind and solar curtailment penalties, overcoming the limitations of traditional methods that only focus on the overall resource capacity. Furthermore, through interval number transformation and interval optimization logic, it effectively handles uncertainties such as market electricity price fluctuations and resource output deviations, ensuring the robustness and economy of the electricity reporting strategy. Therefore, the technical solution of this invention solves the problem of insufficient optimization of peak-shaving market revenue caused by neglecting the characteristics of adjustable loads and the penalties for wind and solar curtailment in existing technologies, improving the adaptability and reliability of the electricity reporting strategy in complex market environments.

[0102] In some embodiments, operational data includes adjustable load resource regulation capacity data, environmental monitoring data, and equipment operating status data; peak-shaving market policy information includes market access power, deviation assessment coefficient, and real-time electricity price; the step of obtaining operational data of load-side distributed resources and peak-shaving market policy information includes:

[0103] Data on regulation capacity and equipment operation status are collected by smart meters and terminal data acquisition devices deployed on the load side.

[0104] Among them, the regulating capacity data refers to the adjustable power capacity of adjustable load resources during peak shaving, in kilowatts; the equipment operation status data includes equipment operation status identifiers (such as running, stopped, fault) and real-time power output values.

[0105] Specifically, smart meters collect load power data once per minute, using Modbus RTU (a serial communication protocol) as the transmission protocol; the terminal acquisition device records the device status through a digital input module and stores it as a time-series database.

[0106] For example, the adjustable load adjustment capacity of a commercial building ranges from 200 kW to 800 kW, and the equipment operation status data is recorded as follows: timestamp 2023-08-01T10:00:00, power 350 kW, status: running.

[0107] Environmental monitoring data is collected through an environmental sensor network deployed on a distributed resource side.

[0108] The environmental monitoring data includes parameters such as temperature, humidity, and light intensity, which are used to assess the impact of the environment on the output of distributed resources.

[0109] Specifically, the environmental sensor network consists of multiple sensor nodes, each equipped with a temperature sensor (measurement range -40 degrees Celsius to 85 degrees Celsius), a humidity sensor (measurement range 0% to 100% relative humidity), and a light sensor (measurement range 0 to 2000 watts per square meter). The data is aggregated to the gateway via the Zigbee protocol (a low-power wireless communication protocol).

[0110] For example, environmental sensors deployed in a photovoltaic power plant collected data showing an irradiance of 1000 watts per square meter, an ambient temperature of 30 degrees Celsius, and a humidity of 60%.

[0111] Market access power, deviation assessment coefficient, and real-time electricity price are obtained through external electricity market data interfaces.

[0112] Among them, market access power refers to the power threshold requirement for participating entities in the peak-shaving market, in kilowatts; deviation assessment coefficient refers to the penalty ratio for the deviation between actual output and declared value, dimensionless; real-time electricity price refers to the current market clearing price, in yuan per kilowatt-hour.

[0113] Specifically, the electricity market data interface uses a RESTful API (Representational State Transfer Application Programming Interface) to interact with the market platform, making requests every 15 minutes and returning data packets in JSON format (a lightweight data exchange format, JavaScript Object Notation).

[0114] For example, data obtained from the market interface shows that the market access power is 500 kW, the deviation assessment coefficient is 0.03, and the real-time electricity price is 0.75 yuan per kilowatt-hour.

[0115] Therefore, according to the above implementation method, the system can realize the real-time collection and integration of operational data and policy information, providing a data foundation for the subsequent construction of a dual-constraint power generation model.

[0116] In some embodiments, a dual-constraint power generation model is constructed based on operational data and peak-shaving market policy information, including:

[0117] The maximum and minimum regulation capacities are extracted from the regulation capacity data to form a set of physical operation characteristic parameters, which are used to characterize the boundary of resource output capacity.

[0118] Among them, the physical operation characteristic parameter set is a data structure used to store the upper and lower limits of the adjustable power of adjustable load resources; for example, the maximum adjustable capacity is extracted as 800 kilowatts and the minimum adjustable capacity is 200 kilowatts from the adjustable capacity data of a commercial building.

[0119] Power upper limit constraints and compliance requirements are extracted from market access power and deviation assessment coefficients, and a market policy constraint parameter set is constructed based on power upper limit constraints and compliance requirements. The market policy constraint parameter set is used to characterize the market compliance boundary.

[0120] The market policy constraint parameter set includes power limits and deviation tolerance parameters defined by market rules. For example, if the market access power is 1000 kilowatts and the deviation assessment coefficient is 0.05 (meaning that the allowed power deviation is 5%), then the power upper limit constraint is 1000 kilowatts and the compliance requirement is that the deviation does not exceed 5%.

[0121] Based on the set of physical operating characteristic parameters and the set of market policy constraint parameters, the range of power that can be declared by the load-side adjustable resources in different time periods is calculated.

[0122] The declared power range refers to the power range that resources can declare under the conditions of physical capacity and market rules, which is solved by a linear programming algorithm (a mathematical optimization method). For example, for the time period from 10:00 to 11:00, the declared power range is calculated to be 300 kilowatts to 500 kilowatts.

[0123] A dynamic threshold adjustment strategy is adopted for compliance verification. The current dynamic threshold adjustment strategy includes verification boundaries determined based on current environmental monitoring data and equipment operating status data.

[0124] Among them, the dynamic threshold adjustment strategy is an adaptive algorithm that dynamically adjusts the verification standard according to real-time environmental conditions and equipment status; the verification boundary refers to the allowable power range fluctuation limit; for example, when the ambient temperature reaches 40 degrees Celsius, the verification boundary tightens, and the declared power range is adjusted to 280 kW to 480 kW; when the equipment operating status shows abnormality, the verification boundary expands, and the allowable range increases by 10%.

[0125] After compliance verification is completed, the dual-constraint power generation model is constructed based on the verification boundaries.

[0126] The dual-constraint power generation model is a mathematical model that integrates physical operating characteristics and market policy constraints, and its output is the final power value that can be declared.

[0127] For example, the model output can declare a power of 350 kW, which simultaneously meets the resource output capacity (200-800 kW) and market compliance requirements (deviation not exceeding 5%).

[0128] Therefore, according to the above implementation method, the system can automatically and accurately determine the range of power that can be declared, ensuring that the declaration of peak shaving market complies with both resource physical limits and market rules, and improving the reliability and economy of the declaration strategy.

[0129] In some embodiments, based on a dual-constraint power generation model, a distributed resource aggregation model is constructed by integrating resource characteristic parameters and wind / solar curtailment penalty parameters, including:

[0130] Extract the resource type proportion weight and resource type identifier from the resource characteristic parameters.

[0131] Among them, the resource type weight refers to the relative importance ratio of different distributed resource types in the aggregation model, which is dimensionless; the resource type identifier refers to the classification code of the resource, which is used to distinguish resource categories, such as adjustable load, photovoltaic, wind power, etc.

[0132] Specifically, the system reads resource characteristic parameters from configuration files or databases, and parses weights and identifiers; the data is stored in JSON format and supports dynamic updates.

[0133] For example, the weighting of resource type is 0.6 for adjustable load and 0.4 for photovoltaic; the resource type identifier is represented by a numerical code, such as 1 for adjustable load and 2 for photovoltaic.

[0134] Extract the unit curtailment penalty price from the wind and solar curtailment penalty parameters.

[0135] Among them, the unit curtailment penalty price refers to the penalty fee payable for each unit of curtailed wind or solar power, expressed in yuan per kilowatt-hour.

[0136] Specifically, the system obtains penalty prices in real time through the electricity market data interface or loads default values ​​from historical data; the verification logic ensures that the price is non-negative and conforms to market rules.

[0137] For example, the unit curtailment penalty price is 0.05 yuan per kilowatt-hour, and the penalty cost when curtailment of solar power occurs is the amount of curtailment multiplied by this price.

[0138] Based on the range of applicable power, resource type weights, resource type identifiers, and unit curtailment penalty prices in the dual-constraint power generation model, an aggregate objective function and aggregate constraints are constructed.

[0139] The aggregate objective function is a mathematical function that aims to minimize the total cost, which includes resource operation costs, wind and solar curtailment penalty costs, and market transaction costs. Aggregate constraints refer to the restrictions that need to be met in the optimization problem, such as power balance and equipment capacity limitations.

[0140] Specifically, the system uses a linear programming algorithm to construct the objective function and constraints; the coefficients of the objective function are calculated by weighting the weights and penalty prices, and the constraints are set based on the range of power that can be declared.

[0141] For example, the aggregation objective function is to minimize the total cost, and its expression is:

[0142] ;

[0143] The constraints include a power range of 300 kW to 500 kW that can be declared, and a power deviation of no more than 5%.

[0144] Based on the aggregation objective function and aggregation constraints, the resource aggregation status and power reporting reference values ​​are calculated.

[0145] Among them, the resource aggregation status indicates which distributed resources have been selected to participate in the aggregation, usually in the form of a binary vector; the electricity declaration reference value refers to the amount of electricity recommended to be declared to the peak-shaving market, in kilowatt-hours.

[0146] Specifically, the system obtains the optimal solution by solving optimization problems (such as using the simplex method); the resource aggregation status is output by decision variables, and the reference value for electricity reporting is derived from the objective function value.

[0147] For example, the resource aggregation state vector is [1,0,1], indicating that adjustable loads and wind power are selected, while photovoltaics are not selected; the reference value for electricity declaration is 400 kWh, corresponding to the time period from 10:00 to 11:00.

[0148] A distributed resource aggregation model is constructed based on the resource aggregation status and the reference value of electricity declaration.

[0149] Among them, the distributed resource aggregation model is a digital model that encapsulates aggregation results, optimization rules, and input / output interfaces to support subsequent decision-making.

[0150] Specifically, the system serializes resource aggregation status, electricity reporting reference values, and related parameters into model files; the model format adopts Protobuf (a language-neutral data serialization format) or JSON to ensure cross-platform compatibility.

[0151] For example, the model output is a file containing a resource list, reported electricity consumption, timestamp, and verification code; model version V1.0 supports dynamic loading and verification.

[0152] Therefore, according to the above implementation method, the system can automatically and efficiently construct a distributed resource aggregation model, realize resource optimization screening and accurate electricity reporting, and improve the profitability and reliability of peak shaving market participation.

[0153] In some embodiments, non-stationary variables are extracted from resource aggregation status and electricity reporting reference values, and the non-stationary variables are converted into interval numbers, including:

[0154] Extract the adjustable load output prediction deviation variable from the resource aggregation state.

[0155] Among them, the adjustable load output prediction deviation variable refers to the deviation between the actual output and the predicted output of adjustable load resources, in kilowatts; this variable is used to quantify the impact of prediction error on dispatching decisions.

[0156] Specifically, the system parses the real-time output data of the selected resource from the binary vector of the resource aggregation status and compares it with the predicted value to calculate the deviation; the data is stored in a time series database and the sampling frequency is once per minute.

[0157] For example, if the predicted output of an adjustable load is 500 kW and the actual output is 480 kW, then the deviation is -20 kW.

[0158] Market electricity price fluctuation variables are extracted from the reference values ​​for electricity declarations.

[0159] Among them, the market electricity price fluctuation variable refers to the real-time fluctuation range of the peak-shaving market electricity price, in yuan per kilowatt-hour; this variable is used to capture market price uncertainty.

[0160] Specifically, the system extracts the electricity price sequence from the historical records of electricity declaration reference values ​​and calculates the difference or standard deviation between adjacent time points as the fluctuation amount; the data processing uses a sliding window algorithm with a window size of 15 minutes.

[0161] For example, the real-time electricity price fluctuated from 0.8 yuan per kilowatt-hour to 0.9 yuan per kilowatt-hour, with a fluctuation range of 0.1 yuan per kilowatt-hour.

[0162] Based on the adjustable load output prediction deviation variable and the market electricity price fluctuation variable, the Latin hypercube sampling method is used to generate multiple sets of sample datasets, which are used to cover the possible value range of non-stationary variables.

[0163] Latin hypercube sampling (LHS) is a statistical sampling method that ensures uniform coverage of the variable space through stratified sampling. The sample dataset refers to a collection of multiple sets of variable values ​​used to simulate uncertain scenarios.

[0164] Specifically, the system uses the Python library SciPy to implement LHS sampling, generating 1000 sets of sample data, each set containing deviation and electricity price fluctuation values; the sampling range is set based on historical data, such as deviation range [-50, 50] kilowatts, and electricity price fluctuation range [0.5, 1.0] yuan per kilowatt-hour.

[0165] For example, the sample dataset includes deviation values ​​of [-10, 20, -5, ...] kilowatts and electricity price fluctuation values ​​of [0.7, 0.8, 0.6, ...] yuan per kilowatt-hour, covering 90% of possible scenarios.

[0166] The upper and lower bounds of the interval numbers for the adjustable load output prediction deviation variable and the market electricity price fluctuation variable are calculated based on the range of possible values ​​for the unsteady variable.

[0167] Among them, the upper and lower bounds of the interval number refer to the minimum and maximum possible values ​​of the variable, forming a closed interval [lower bound, upper bound], which is used to encapsulate uncertainty; the calculation method is to use the percentiles or extreme values ​​of the statistical sample dataset.

[0168] Specifically, the system calculates the 5% and 95% quantiles of the sample data as upper and lower bounds to avoid the influence of outliers; for example, if the 5% quantile of the deviation is -30 kW and the 95% quantile is 30 kW, then the interval is [-30, 30] kW.

[0169] For example, for the variable of electricity price fluctuation, the range of [0.6, 0.95] yuan per kilowatt-hour is calculated.

[0170] The upper and lower bounds of the interval number are updated using a preset dynamic correction strategy.

[0171] Among them, the dynamic correction strategy is an adaptive algorithm that iteratively adjusts the interval boundary based on real-time data to reflect the dynamic changes of the system; the preset refers to the initialization of the algorithm parameters to default values, such as a correction period of 15 minutes.

[0172] Specifically, the system uses a sliding window algorithm (window size 24 hours) to compare historical actual values ​​with predicted values. If the deviation continues to exceed the current range, the boundary is expanded proportionally. For example, if the actual deviation exceeds the range three times in a row, the upper and lower bounds are expanded by 10%.

[0173] For example, the initial range is [-30, 30] kW, and the updated range is [-33, 33] kW because the actual deviation reaches 35 kW.

[0174] Output the transformed interval number form based on the updated upper and lower bounds of the interval number.

[0175] The interval number form is a mathematical representation that encapsulates variables as interval objects, supporting interval operations and optimizations; the output format is a standardized data structure, such as a JSON (a lightweight data exchange format) file.

[0176] Specifically, the system serializes the upper and lower bounds of the interval into key-value pairs and adds timestamps and confidence metadata; for example, the output content is {"variable type":"deviation","lower bound":-33,"upper bound":33,"unit":"kilowatt"}.

[0177] For example, the interval number form is used as the input for the subsequent interval optimization module.

[0178] Therefore, according to the above implementation method, the system can effectively capture and process the uncertainty in the peak shaving market, transform non-stationary variables into computable interval numbers, provide robust input for optimized scheduling, and improve the adaptability and reliability of the reporting strategy.

[0179] In some embodiments, the optimal electricity reporting strategy is obtained by solving interval optimization logic based on the transformed non-steady variables, including:

[0180] Based on non-stationary variables in the form of interval numbers, construct an interval number objective function.

[0181] Among them, the interval number objective function is a mathematical function whose input is an interval number variable (such as the adjustable load output prediction deviation interval [-30, 30] kW and the market electricity price fluctuation interval [0.6, 0.95] yuan per kilowatt-hour), and the output is the possible range of the objective value. It is used to optimize decision-making under uncertainty. The function is usually in the form of minimizing or maximizing interval cost or benefit.

[0182] Specifically, the system uses interval arithmetic rules to construct functions, for example:

[0183] Total cost range = operating cost × weight + penalty cost × range penalty price;

[0184] Weights and cost parameters are loaded from historical data.

[0185] For example, the objective function of the interval number is expressed as minimizing the total cost, with an interval of [1000, 1500] yuan, corresponding to the range of electricity that can be declared.

[0186] A pre-defined interval sorting relationship comparison method is used to perform comparison calculations on the objective function of interval numbers to obtain the objective trade-off coefficients.

[0187] Among them, the interval sorting relationship comparison method is an algorithm used to compare the size or priority of different interval numbers, thereby quantifying the trade-off between the expected value and stability of the objective function; the objective trade-off coefficient is a scalar value (ranging from 0 to 1) used to balance risk and reward in optimization.

[0188] Specifically, the system uses the probability degree method (an interval comparison method) to calculate the coefficient. For example, it compares the probability degree of cost intervals [1000, 1500] yuan with [1200, 1400] yuan, and the output coefficient of 0.7 indicates that more emphasis is placed on stability.

[0189] For example, the target tradeoff coefficient is set to 0.7, which means that 70% of the weight of the optimization process considers the worst-case cost.

[0190] Based on the objective trade-off coefficient, the objective function of the interval number is iteratively optimized using the current reinforcement learning algorithm, which includes optimization parameters that are updated following current market data and historical strategy data.

[0191] Among them, reinforcement learning algorithms are a type of machine learning method that learns the optimal strategy through the interaction between an agent and its environment; optimization parameters include learning rate, discount factor, etc., which are dynamically adjusted based on real-time market data (such as electricity price fluctuations) and historical strategy performance.

[0192] Specifically, the system uses the Q-learning algorithm (a value-iterative reinforcement learning algorithm), with the learning rate initialized to 0.1 and the discount factor set to 0.9. The parameters are updated every 15 minutes, and the Q-value (state-action value) is recalculated based on the latest market data.

[0193] For example, the optimization process iterates 100 times, and each update uses the current electricity price data (e.g., 0.8 yuan per kilowatt-hour) to adjust the strategy to minimize the interval cost.

[0194] When the preset convergence condition is met, the iterative optimization process of the interval number objective function is terminated, and the current solution of the interval number objective function is output.

[0195] Among them, the convergence condition refers to the criterion for stopping the optimization process, which is usually based on the rate of change of the objective function or the number of iterations; the interval number objective function solution is the optimized interval result, which represents the range of the optimal objective value.

[0196] Specifically, the system is set to convergence condition that the objective function changes by less than 1% in 10 consecutive iterations or the maximum number of iterations is 200; the solution output is in interval form, such as the cost interval [1050, 1450] yuan.

[0197] For example, when the cost range stabilizes from [1100, 1500] yuan to [1050, 1450] yuan, the rate of change is less than 1%, the optimization is terminated and the solution is output.

[0198] The objective function of the interval number is demapped into the power reporting strategy space to generate the optimal power reporting strategy, which includes the reported power value and the reporting time series.

[0199] The mapping process transforms mathematical interval solutions into practically executable reporting strategies; the electricity reporting strategy space refers to the set of all possible reporting actions, including electricity values ​​and time points; the unit of the reported electricity value is kilowatt-hours, and the reporting time sequence is a list of time points.

[0200] Specifically, the system uses a linear interpolation algorithm to demap the cost range to the declared electricity range, such as a cost range of [1050, 1450] yuan corresponding to a declared electricity range of [300, 500] kilowatt-hours; the time series is generated based on the market clearing cycle, such as a time point every 15 minutes.

[0201] For example, the generated optimal electricity reporting strategy includes reporting 400 kWh of electricity (taking the midpoint of the interval), with a time series of reporting 200 kWh at 10:00 and reporting 200 kWh at 10:15.

[0202] Therefore, according to the above implementation method, the system can automatically and efficiently solve the optimal electricity reporting strategy, handle uncertainty through interval optimization, improve the economy and robustness of the reporting strategy, and ensure stable returns in the peak shaving market.

[0203] Figure 4 This is a structural block diagram of a distributed resource peak-shaving market electricity reporting system according to an embodiment of the present invention.

[0204] like Figure 4 As shown, the peak-shaving market electricity reporting system for this distributed resource includes:

[0205] The operation and market data acquisition module 210 is used to acquire operation data of distributed resources on the load side to obtain peak shaving market policy information.

[0206] The power generation model construction module 220 is used to construct a dual-constraint power generation model based on operating data and peak-shaving market policy information.

[0207] The resource aggregation model construction module 230 is used to construct a distributed resource aggregation model based on the dual-constraint power generation model, integrating resource characteristic parameters and wind and solar curtailment penalty parameters. The distributed resource aggregation model is used to generate resource aggregation status and power declaration reference values.

[0208] The unsteady variable extraction module 240 is used to extract unsteady variables from the resource aggregation status and the reference value of electricity declaration, and to convert the unsteady variables into the form of interval numbers.

[0209] The electricity declaration strategy generation module 250 is used to obtain the optimal electricity declaration strategy by using interval optimization logic based on the transformed non-steady variables, so as to generate an electricity declaration scheme for participating in the peak shaving market.

[0210] The specific functions and examples of each module and submodule of the device in this embodiment of the invention can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0211] According to embodiments of the present invention, the above-described method of the present invention can be applied to an electronic device and a readable storage medium.

[0212] Figure 5 A schematic block diagram of an example electronic device 600 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0213] like Figure 5 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0214] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and communication unit 609, such as network card, modem, wireless transceiver, etc. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0215] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as a method for reporting peak-shaving market electricity demand for distributed resources. For example, in some embodiments, a method for reporting peak-shaving market electricity demand for distributed resources can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the method for reporting peak-shaving market electricity demand for distributed resources described above can be performed. Alternatively, in other embodiments, the computing unit 601 may be configured, by any other suitable means (e.g., by means of firmware), to perform a distributed resource peak-shaving market electricity reporting method.

[0216] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0217] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0218] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0219] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT or LCD monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual, auditory, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0220] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0221] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0222] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0223] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for reporting peak-shaving electricity in a distributed resource market, characterized in that, include: Obtain operational data of distributed resources on the load side to obtain information on peak-shaving market policies; A dual-constraint power generation model is constructed based on the aforementioned operational data and peak-shaving market policy information; Based on the dual-constraint power generation model, resource characteristic parameters and wind and solar curtailment penalty parameters are integrated to construct a distributed resource aggregation model, which is used to generate resource aggregation status and power declaration reference values. Unsteady variables are extracted from the resource aggregation status and electricity declaration reference value, and the unsteady variables are converted into interval numbers. Based on the transformed unsteady variables, the optimal electricity declaration strategy is obtained by using interval optimization logic to generate an electricity declaration scheme for participating in the peak shaving market. The steps for constructing a distributed resource aggregation model include: Extract the resource type proportion weight and resource type identifier from the resource characteristic parameters; Extract the unit curtailment penalty price from the aforementioned wind and solar curtailment penalty parameters; Based on the declared power range, the resource type weight, the resource type identifier, and the unit curtailment penalty price in the dual-constraint power generation model, an aggregate objective function and aggregate constraints are constructed. Based on the aggregation objective function and the aggregation constraints, the resource aggregation status and the reference value for electricity declaration are calculated. The distributed resource aggregation model is constructed based on the resource aggregation status and the electricity reporting reference value.

2. The method according to claim 1, characterized in that, The operational data includes adjustable load resource regulation capacity data, environmental monitoring data, and equipment operating status data; the peak-shaving market policy information includes market access power, deviation assessment coefficient, and real-time electricity price; the steps of obtaining operational data of load-side distributed resources and peak-shaving market policy information include: The regulating capacity data and the equipment operating status data are collected by smart meters and terminal acquisition devices deployed on the load side; The environmental monitoring data is collected through an environmental sensor network deployed on a distributed resource side; The market access power, the deviation assessment coefficient, and the real-time electricity price are obtained through an external electricity market data interface.

3. The method according to claim 2, characterized in that, The step of constructing a dual-constraint power generation model based on the operational data and peak-shaving market policy information includes: The maximum and minimum regulation capacities are extracted from the regulation capacity data to form a set of physical operation characteristic parameters, which are used to characterize the boundary of resource output capacity. Power upper limit constraints and compliance requirements are extracted from the market access power and the deviation assessment coefficient, and a market policy constraint parameter set is constructed based on the power upper limit constraints and compliance requirements. The market policy constraint parameter set is used to characterize the market compliance boundary. Based on the set of physical operating characteristic parameters and the set of market policy constraint parameters, the range of power that can be declared for the load-side adjustable resources in different time periods is calculated. The current dynamic threshold adjustment strategy is used to verify the compliance of the declared power range. The current dynamic threshold adjustment strategy includes a verification boundary determined based on the current environmental monitoring data and equipment operating status data. After the compliance verification is completed, the dual-constraint power generation model is constructed based on the verification boundary.

4. The method according to claim 3, characterized in that, The step of extracting non-stationary variables from the resource aggregation status and electricity declaration reference values, and converting the non-stationary variables into interval numbers, includes: Extract adjustable load output prediction deviation variables from the resource aggregation state; Extract market electricity price fluctuation variables from the aforementioned electricity declaration reference values; Based on the adjustable load output prediction deviation variable and the market electricity price fluctuation variable, multiple sets of sample datasets are generated using the Latin hypercube sampling method. These multiple sets of sample datasets are used to cover the possible value range of the non-stationary variable. The upper and lower bounds of the interval numbers for the adjustable load output prediction deviation variable and the market electricity price fluctuation variable are calculated based on the range of possible values ​​of the unsteady variable. The upper and lower bounds of the interval number are updated using a preset dynamic correction strategy; Output the transformed interval number form based on the updated upper and lower bounds of the interval number.

5. The method according to claim 4, characterized in that, The step of obtaining the optimal electricity reporting strategy by using interval optimization logic based on the transformed non-steady variables includes: Based on the non-stationary variables in the form of interval numbers, construct an interval number objective function; A preset interval sorting relationship comparison method is used to perform comparison calculation processing on the objective function of the interval numbers to obtain the objective trade-off coefficient; Based on the target trade-off coefficient, the objective function of the interval number is iteratively optimized using the current reinforcement learning algorithm, which includes optimization parameters that are updated following current market data and historical strategy data. When the preset convergence condition is met, the iterative optimization process of the interval number objective function is terminated, and the current solution of the interval number objective function is output. The objective function of the interval number is demapped into the power reporting strategy space to generate the optimal power reporting strategy, which includes the reported power value and the reporting time series.

6. The method according to claim 5, characterized in that, The steps for generating an electricity declaration plan to participate in the peak-shaving market include: Cross-regional power balance verification was performed on the declared electricity values ​​and declaration time series in the optimal electricity declaration strategy. Based on the results of the cross-regional power balance verification, the declared electricity value and the declaration time series are updated to obtain a preliminary electricity declaration plan. The feasibility of the preliminary electricity declaration plan is verified based on the deviation assessment coefficient and the market access power. Based on the feasibility verification results of the preliminary electricity declaration plan, the preliminary electricity declaration plan is revised, and the electricity declaration plan is generated. The electricity declaration plan includes the final declared electricity volume, the declaration time node, and related constraints.

7. A distributed resource peak-shaving market electricity reporting system, characterized in that, include: The operation and market data acquisition module is used to acquire operation data of distributed resources on the load side to obtain information on peak shaving market policies. The power generation model construction module is used to construct a dual-constraint power generation model based on the operating data and peak-shaving market policy information. The resource aggregation model construction module is used to construct a distributed resource aggregation model based on the dual-constraint power generation model, integrating resource characteristic parameters and wind and solar curtailment penalty parameters. The distributed resource aggregation model is used to generate resource aggregation status and power declaration reference values. The unsteady variable extraction module is used to extract unsteady variables from the resource aggregation status and the electricity declaration reference value, and convert the unsteady variables into interval numbers. The electricity declaration strategy generation module is used to obtain the optimal electricity declaration strategy by using interval optimization logic based on the transformed unsteady variables, so as to generate an electricity declaration scheme for participating in the peak shaving market. The resource aggregation model construction module is also used to extract the resource type proportion weight and resource type identifier from the resource characteristic parameters; Extract the unit curtailment penalty price from the aforementioned wind and solar curtailment penalty parameters; Based on the declared power range, the resource type weight, the resource type identifier, and the unit curtailment penalty price in the dual-constraint power generation model, an aggregate objective function and aggregate constraints are constructed. Based on the aggregation objective function and the aggregation constraints, the resource aggregation status and the reference value for electricity declaration are calculated. The distributed resource aggregation model is constructed based on the resource aggregation status and the electricity reporting reference value.

8. An electronic device, characterized in that, include: At least one processor; as well as The memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, Computer instructions are used to cause a computer to perform the method according to any one of claims 1-6.

Images