A tea can coordinate distributed energy system scheduling method

Abstract

The application provides a tea energy cooperative distributed energy system scheduling method, comprising: establishing an electric power demand model and a thermal power demand model for a tea production process series production process chain, and correspondingly setting process timing constraints; constructing a mixed integer linear programming optimization model, the decision variables of which include process scheduling variables, electric side energy supply variables and thermal side energy supply variables; performing day-ahead global solution and intra-day rolling optimization on the mixed integer linear programming optimization model to output scheduling execution instructions for the current period. The application realizes systematic and unified description of the electric and thermal load characteristics of the whole production process by establishing a demand model, which can support overall analysis of the multi-energy operation characteristics of the whole production process; by including the process scheduling variables and the electric side and thermal side energy supply variables in the same optimization model, the operation optimization space of the multi-energy complementary system is effectively expanded; by adopting day-ahead and intra-day architecture optimization, the energy system operation stability is considered, the execution deviation is reduced, and the landing feasibility of the scheduling scheme is improved.

Description

Technical Field

[0001] This application relates to the field of energy dispatching technology, specifically to a tea-energy collaborative distributed energy system dispatching method. Background Technology

[0002] Existing energy consumption studies on industrial production processes in tea making mostly focus on individual equipment and processes, analyzing the energy consumption patterns and equipment operating characteristics of a single link. However, for a complete production chain consisting of multiple processes connected in series, there is a lack of systematic and unified description of the temporal connection relationship of energy consumption in each process and the coupling distribution characteristics of electrical and thermal loads, making it difficult to directly support a holistic analysis of the multi-energy operation characteristics of the entire production process.

[0003] Therefore, in the scheduling optimization of distributed energy systems, existing research mostly takes the energy supply side as the core of optimization, while the energy demand on the production side is usually treated as a pre-given fixed boundary condition, with production arrangements and energy scheduling being decided independently. This approach cannot reflect the dynamic impact of changes in the production process on energy demand, making it difficult to match the optimization scheme of the energy system with actual production, thus limiting the operational optimization space of multi-energy complementary systems.

[0004] Furthermore, based on the above, the existing operation and scheduling schemes take the day-ahead static plan as the actual execution scheme. When faced with uncertainties such as fluctuations in renewable energy output and load forecast deviations, dynamic adjustments are limited to the output allocation of energy equipment. It is difficult to simultaneously take into account the execution rules of production processes and the operational stability of the energy system, which can easily lead to large deviations between the plan and the actual execution and insufficient feasibility of the scheduling scheme.

[0005] In summary, existing technologies cannot simultaneously guarantee the coordinated and optimized scheduling of electrical and thermal loads in distributed energy systems, meet the requirements for cost reduction and low carbon emissions in system operation, and ensure the stability of production execution, while adhering to the operating rules of serial production processes. Summary of the Invention

[0006] This application addresses the problems existing in the prior art by providing a tea-energy collaborative distributed energy system scheduling method that takes into account collaborative optimization scheduling, reduces system operating costs, and meets the requirements of low carbon emissions and production execution stability.

[0007] To achieve the above objectives, the technical solution adopted in this application is as follows: This application provides a tea-energy coordinated distributed energy system scheduling method, which includes: For the tea-making process with combined electric and thermal loads, a series production process chain is established, and an electric power demand model and a thermal power demand model are established for each production process. The process timing constraints of the series production process chain are set accordingly. Based on the electric power demand model and the thermal power demand model, a mixed-integer linear programming optimization model is constructed; the decision variables of the mixed-integer linear programming optimization model include process scheduling variables, electric power supply variables, and thermal power supply variables. The mixed-integer linear programming optimization model is solved globally one day in advance to obtain the baseline scheduling plan; The baseline scheduling plan is subjected to intraday rolling optimization at a preset step size, and the process timing constraints are applied in the rolling time domain to output the scheduling execution instruction for the current time period.

[0008] Optionally, the process timing constraints include single-start constraints, continuous operation constraints, sequential connection constraints between adjacent processes, and allowed start time window constraints for processes.

[0009] Optionally, in establishing the electrical power demand model and thermal power demand model for each production process, the operating status of each process can be used as the input for model calculation. The process scheduling variables are used to obtain the running status of each process through time-series mapping.

[0010] Optionally, the load-side input of the mixed-integer linear programming optimization model includes the total electrical load and total thermal load of the tea-making process's serial production chain; The total electrical load is obtained by summing up the electrical power requirements of each process. The total heat load is obtained by summing up the heat power requirements of each process.

[0011] Optionally, the constraints of the mixed-integer linear programming optimization model may also include electrical power balance constraints, thermal power balance constraints, electrothermal conversion constraints, and equipment operation boundary constraints.

[0012] Optionally, the mixed-integer linear programming optimization model uses a preset weighted multi-objective function as the objective function; The optimization objectives of the weighted multi-objective function include one or a combination of operating cost, carbon emission, net load peak, power purchase fluctuation, and process completion time.

[0013] Optionally, the power supply variables on the electricity side include grid power variables, battery energy storage charging and discharging power variables, renewable energy consumption power variables, and electric heating input power variables.

[0014] Optionally, the heat-side energy supply variables include biomass heating power variables, thermal storage charging and discharging power variables, and electric heating output power variables; Based on the electrothermal conversion efficiency, the electric heating output power is calculated according to the electric heating input power of the series production process chain in the tea-making process.

[0015] Optionally, the benchmark scheduling plan includes a benchmark process scheduling scheme and a benchmark energy supply scheme; The intraday rolling optimization is performed according to a preset time step, and each re-optimization is completed within a preset prediction window to output the scheduling execution instruction for the current time period. The scheduling plan obtained by the intraday rolling optimization is updated with the next rolling cycle.

[0016] Optionally, the intraday rolling optimization is performed based on real-time state deviations to perform re-optimization; In the intraday rolling optimization, the baseline process scheduling scheme and baseline energy supply scheme for subsequent periods are simultaneously modified within or after the current rolling cycle.

[0017] Compared with the prior art, this application has the following advantages: This application establishes separate power demand models for the sequential production processes in tea making, achieving a systematic and unified description of the electrical and thermal load characteristics of the entire production process, which can support the overall analysis of the multi-energy operation characteristics of the entire process. Furthermore, by incorporating process scheduling variables and power supply variables from the electricity and heat sides into the same mixed-integer linear programming optimization model, the operational optimization space of the multi-energy complementary system is effectively expanded. Moreover, by adopting day-ahead and intraday architecture optimization, the application balances the production process execution rules with the stability of energy system operation, reduces the deviation between planned and actual execution, and improves the feasibility of scheduling schemes. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0022] It is worth noting that, unless otherwise specified, the methods used in this application are all conventional methods; and the raw materials and equipment used are all conventional commercially available products, and their sources are not specifically limited.

[0023] It should also be noted that, for ease of understanding, the method steps in the specific embodiments of this application are described in a certain order, but those skilled in the art can change the order of the steps according to actual needs, so this should not be used as a limiting condition; further, in the description of the following specific embodiments, the superscripts and subscripts of each parameter should be understood as distinguishing marks of similar identifiers in accordance with common interpretations, representing the parameters of the related or corresponding devices, and should not be understood as specific models or special marks.

[0024] like Figure 1 As shown, this application provides a tea-energy collaborative distributed energy system scheduling method to achieve coordinated and optimized operation of the tea-making process's serial production steps and multi-energy supply system.

[0025] This embodiment uses a production process chain with a combined electrothermal load, consisting of five sequentially connected processes in tea processing: withering, shaking, fixing, rolling, and drying. The scheduling cycle is 24 hours with a time step of 1 hour. The system comprises two parts: an energy supply side and a production process side. The energy supply side includes a photovoltaic power generation unit, a grid interaction unit, a battery energy storage unit, a biomass heat source unit, an electric heating unit, and a thermal storage unit. On the process side, each process proceeds sequentially in a fixed order, with material output gradually transferred along the process. Energy input is provided to the corresponding process in the form of electricity, heat, or a combination of electricity and heat. The electric heating unit handles the energy conversion between the electric and heat sides, the thermal storage unit buffers the temporal fluctuations of the heat load, and the battery energy storage unit is used for peak shaving and valley filling, as well as renewable energy consumption.

[0026] First, power demand models and thermal power demand models for each production process are established, and corresponding time constraints are set for the sequential production process chain in the tea-making process. In establishing these models, the operating status of each process is used as the input for model calculation. The process scheduling variables are mapped to the operating status of each process through time sequence mapping. The five processes—withering, shaking, fixing, rolling, and drying—are denoted by the subscripts wit, rock, fix, roll, and dry, respectively. The power demand for each process is obtained by multiplying the corresponding process's operating status variable by the power parameter per unit time period. The power demand for the withering process is expressed as: ; In the formula, This represents the electrical power requirement of the withering process during time period t. For the withering process, This refers to the electrical power parameters per unit time period during the withering process.

[0027] The power requirement for the shaking process is expressed as follows: ; In the formula, The electrical power requirement for the shaking process during time period t. For the operating status variables of the shaking and greening process, This refers to the electrical power parameters per unit time period during the shaking process.

[0028] The power requirements for the blanching process are expressed as follows: ; In the formula, The electrical power requirement for the blanching process during time period t. This is a variable representing the operational status of the blanching process. This refers to the electrical power parameters per unit time period during the blanching process.

[0029] The power requirement for the kneading process is expressed as follows: ; In the formula, The electrical power requirement for the kneading process during time period t. For the kneading process, the operating status variable, This refers to the electrical power parameters per unit time period during the kneading process.

[0030] The power requirement for the drying process is expressed as follows: ; In the formula, The electrical power requirement for the drying process during time period t. For the operating status variables of the drying process, This represents the electrical power parameter per unit time period for the drying process.

[0031] The above formula maps the operating status of each process to its power demand, with the main differences being in the type of equipment and the power values. Withering mainly corresponds to the power consumption of fans and auxiliary equipment, shaking and rolling mainly correspond to the power consumption of motor drives, and fixing and drying simultaneously include power requirements for driving, conveying, ventilation, or control.

[0032] The heat power demand is mainly concentrated in the withering, fixation, and drying stages; the shaking and rolling processes do not have independent direct heat loads. The heat power demand for the withering process is expressed as follows: ; In the formula, This indicates the heat power requirement of the withering process during time period t. For the withering process, This refers to the thermal power parameters per unit time period during the withering process.

[0033] The direct heat power demand expression for the shaking process is: ; In the formula, Let t be the direct heat power requirement of the shaking process during time period t; the direct heat power requirement of the shaking process is approximately 0.

[0034] The heat power requirement for the blanching process is expressed as follows: ; In the formula, The heat power requirement for the blanching process during time period t, This is a variable representing the operational status of the blanching process. This refers to the thermal power parameter per unit time period during the blanching process.

[0035] The direct heat power requirement expression for the kneading process is: ; In the formula, This represents the direct heat power requirement of the kneading process during time period t; the direct heat power requirement of the kneading process is approximately 0.

[0036] The heat power requirement for the drying process is expressed as follows: ; In the formula, The heat power requirement of the drying process during time period t. For the operating status variables of the drying process, This represents the thermal power parameter per unit time period of the drying process.

[0037] Of the five processes mentioned above, withering, fixation, and drying correspond to explicit heat loads, while shaking and rolling do not have independent heat loads. This treatment concentrates the heat load on processes that have a more significant impact on quality control and moisture content changes.

[0038] The total electrical load of the tea-making process is obtained by superimposing the electrical power requirements of each process, and the total thermal load of the tea-making process is obtained by superimposing the thermal power requirements of each process. The total electrical load and total thermal load are used as load-side inputs for the subsequent optimization model and participate in the calculation of electrical power balance and thermal power balance, respectively.

[0039] Therefore, the total electrical load of the tea-making process can be expressed as: ; The total heat load of the tea-making process is expressed as: ; In the formula, and These represent the total electrical load and total heat load of tea production during time period t, respectively.

[0040] Process timing constraints include single-start constraints, continuous operation constraints, adjacent process sequence constraints, and allowed start time window constraints. These constraints are used to define the execution rules of the production process and ensure that process operation meets process requirements. The start and operation variables of each process in discrete time are defined as binary variables, and the process set is... The time set is The starting variable for process s is The runtime variable is The single-start constraint for each process requires that each process start exactly once within the scheduling cycle, expressed as: ; In the formula, The starting variable for process s in time period t is used to ensure that each process starts exactly once within the scheduling cycle, thus avoiding the model obtaining a low-cost solution that does not conform to the production logic by starting multiple times in segments.

[0041] The continuous operation constraint of the process is implemented through the start-run convolution relationship, that is, the process operation state at a certain time period is equal to the sum of all start variables within that time period and the duration of the corresponding process before it. This is used to handle the logic of continuous operation after start-up, realizing the analytical mapping from discrete start point to process load curve, expressed as: ; In the formula, This represents the operating status of process s during time period t. Let be the duration of process s. This formula is used to handle the logic of continuous operation after startup and to implement the analytical mapping from discrete startup points to process load curves.

[0042] Define a process start time variable to compress the binary start decision into a single time variable, which is used to uniformly express the process connection relationship. The start time is expressed as: ; In the formula, This represents the discrete start time of process s.

[0043] Each process has a corresponding allowed start time window, which limits the upper and lower boundaries of the process start time. Let the allowed start window of process s be denoted as . , This represents the lower bound of the allowed start time for the process. This represents the upper bound of the allowed start time for a process. The sequential connection constraint between adjacent processes includes minimum and maximum waiting constraints. The start time variable limits the range of the difference between the start time of the preceding and following processes. The minimum waiting constraint between adjacent processes is expressed as: ; In the formula, This is the minimum waiting time between adjacent processes, used to ensure that necessary process connection time is reserved after the previous process is completed.

[0044] The maximum waiting constraint between adjacent processes is expressed as: ; In the formula, This is the maximum waiting time between adjacent processes, used to avoid excessively long intervals between processes that disrupt the process cycle.

[0045] In summary, single-start, continuous operation, and sequential waiting constraints together constitute the basic conditions for the feasibility of tea-making processes. Specifically, the single-start constraint ensures that each process is executed only once, the continuous operation constraint prevents processes from being split and executed, and the sequential waiting constraint ensures that preceding and subsequent processes are connected according to the process rhythm. These process constraints also determine the timing of electrical and heating loads and interact with energy-side constraints such as energy storage, heating, and electricity purchase and sale. Therefore, the scheduling problem in the tea-making scenario cannot be equated with general transferable load scheduling; it requires a combined approach of process scheduling and energy supply and demand balance within the same model.

[0046] The above steps in this embodiment construct an electrothermal time-series load model oriented towards the constraints of the entire tea-making process. The material input-output relationship of processes such as withering, shaking, fixing, rolling and drying is transformed into the time-series expression of electrical power and thermal power of each process, clarifying the dynamic characteristics of the electrothermal load formed by the tea-making process chain. By introducing constraints such as process time windows, continuous operation, sequential connection and equipment status, the simplification of tea processing to a general transferable load is avoided, and a rigorous mathematical characterization of process executability is achieved.

[0047] Based on the power demand model and the heat demand model, a mixed-integer linear programming (MILP) optimization model is constructed. The decision variables of the MILP optimization model include three categories: process scheduling variables, electricity supply variables, and heat supply variables. The electricity supply variables include grid power variables, battery storage charging and discharging power variables, renewable energy consumption power variables, and electric heating input power variables. In this embodiment, renewable energy consumption power corresponds to photovoltaic power consumption power, and grid power variables include grid-purchased power and grid-connected power. The heat supply variables include biomass heating power variables, thermal storage charging and discharging power variables, and electric heating output power variables. Based on the electrothermal conversion efficiency, the electric heating output power is calculated according to the electric heating input power of the tea-making process's series production process chain, thereby realizing the coupling relationship between electricity and heat energy.

[0048] Furthermore, the constraints of the mixed-integer linear programming optimization model also include electrical power balance constraints, thermal power balance constraints, electrothermal conversion constraints, and equipment operation boundary constraints. All feasible solutions obtained by solving the model must simultaneously satisfy all constraints.

[0049] The power balance constraint limits the algebraic sum of purchased power, grid-connected power, battery discharge power, battery charging power, and local photovoltaic utilization power in each time period, maintaining a supply-demand balance with the base load, production process load, and electric heating power consumption, expressed as: ; In the formula, and These are the purchased power and the grid-connected power, respectively. and These are the battery discharge power and charging power, respectively. To utilize photovoltaic power locally, As basic electrical load, For the electrical load of tea production, This refers to the power consumption of electric heating.

[0050] The thermal power balance constraint limits the algebraic sum of biomass heating power, electric heating output power, thermal storage heat release power, and thermal storage heat charging power in each time period, maintaining a supply-demand balance with the base heat load and the production process heat load, expressed as: ; In the formula, For biomass heating power, For electrothermal conversion efficiency, and These are the heat storage and heat release power and the heat charging power, respectively. Based on the base heat load.

[0051] The electrothermal conversion constraint stipulates that the output power of the electric heater is equal to the product of the input power and the electrothermal conversion efficiency, expressed as: ; In the formula, It provides electric heating output.

[0052] Battery state update constraints characterize the energy accumulation and release process of a battery over a continuous period of time, and are expressed as: ; In the formula, Let be the battery state quantity for time period t. and These are charging efficiency and discharging efficiency, respectively. The duration is the unit of time period.

[0053] The thermal storage state update constraint characterizes the state evolution of the thermal buffer capacity in time-series scheduling, and is expressed as: ; In the formula, Let be the thermal storage state quantity for time period t. and These are the thermal storage charging efficiency and the thermal storage releasing efficiency, respectively.

[0054] Equipment power boundary constraints encompass the upper and lower limits of grid interaction power and the upper limit of usable photovoltaic power, expressed as: ; In the formula, and These are the upper limits for purchased electricity and the upper limits for grid connection, respectively. Let be the photovoltaic power available during time period t.

[0055] The charge-discharge mutual exclusion constraint ensures, through binary state variables, that the energy storage device operates in only one direction at a time. The battery charge-discharge mutual exclusion constraint is expressed as: ; In the formula, The battery charge / discharge state is a binary variable. and These are the upper limits for charging power and discharging power, respectively.

[0056] The mutual exclusion constraint of thermal storage charging and discharging is expressed as: ; In the formula, For the thermal storage charging and discharging state, a binary variable, and These are the upper limits for thermal storage and charging power and the upper limit for heat release power, respectively.

[0057] State boundaries and initial state constraints are used to ensure that the energy storage device always operates within a safe and available range, and the initial conditions for the state update equations are given, expressed as: ; In the formula, , , , For the corresponding state boundary, and These are the initial state quantities for the battery and thermal storage, respectively.

[0058] The mixed-integer linear programming optimization model uses a pre-defined weighted multi-objective function as its objective function. The optimization objectives of this weighted multi-objective function include one or more of the following: operating cost, carbon emissions, peak net load, power purchase fluctuation, and process completion time. The appropriate optimization objective can be selected and weighted according to actual operational needs. The overall weighted multi-objective function is expressed as follows: ; In the formula, , , , and These represent operating costs, carbon emissions, peak net load, fluctuations in purchased electricity, and process completion time targets, respectively. , , and These are the corresponding weighting coefficients.

[0059] The operating cost item is calculated jointly from the cost of electricity purchase, revenue from electricity sales, and biomass heating costs, and is expressed as follows: ; In the formula, and These represent the unit cost of purchasing electricity and the unit cost of selling electricity, respectively, for time period t. The unit cost coefficient for biomass heating.

[0060] The carbon emissions item is obtained by adding the carbon emissions from electricity purchases from the grid and the carbon emissions from biomass heating, and is expressed as: ; In the formula, and These are the electricity purchase coefficient from the power grid and the carbon emission coefficient from biomass heating, respectively.

[0061] Net load is defined as the difference between the purchased power and the grid-connected power during a given period, expressed as: ; In the formula, This represents the net load during time period t. Power purchased for the power grid This refers to the power consumption for internet access.

[0062] The peak net load term, defined by an auxiliary variable, limits the maximum net power purchase within the dispatching period, representing the maximum supply pressure the system places on the upstream power grid. An auxiliary variable for the upper bound of the maximum net power purchase is introduced. Net load peak target Expressed as: ; The corresponding upper limit constraint for peak net load is expressed as: ; In the formula, To ensure that this constraint holds true for all time periods within the scheduling cycle, this formula guarantees that the net load in any time period will not exceed the auxiliary variable. This allows for the characterization of the peak net load during the scheduling cycle.

[0063] The power purchase fluctuation term characterizes the impact of dispatching results on power fluctuations in the upstream power grid by summing the absolute values ​​of power purchase changes in adjacent time periods. Positive and negative auxiliary variables are introduced. and The change in power purchases between adjacent time periods can be decomposed and expressed as follows: ; The target for fluctuations in purchased electricity power is expressed as: ; In the formula, The target for power purchase fluctuation is T, which represents the number of dispatching cycle periods.

[0064] The process completion time parameter uses the start time of the last process step to characterize the completion time of the entire process chain, preventing the optimization results from excessively delaying subsequent processes. It is expressed as: ; In the formula, This represents the starting variable for the final process in time period t.

[0065] This embodiment establishes a mixed integer linear programming model for tea-making process-energy coordinated scheduling that takes into account multi-energy complementarity. It characterizes the energy coupling relationship between photovoltaics, power grid, battery energy storage, biomass heating, electric heating and thermal storage in the same optimization framework. It incorporates process scheduling and electric and heat supply schemes into joint optimization, and achieves a balance between electric and heat supply and demand under the premise of meeting process constraints and equipment operation limitations.

[0066] In the day-ahead global optimization phase, the day-ahead forecast and equipment parameters are first read to establish a full-time domain MILP model, including process sequence, equipment boundaries, and electrothermal balance constraints. Then, the day-ahead optimization problem is solved to obtain the day-ahead baseline plan, including process start times, equipment output, and energy storage status. The mixed-integer linear programming optimization model is then solved globally for the day-ahead period. Before solving, the nonlinear terms in the model are linearized. The absolute value of power change is decomposed into a linear expression using non-negative auxiliary variables. Logical constraints such as charge-discharge mutual exclusion are transformed into linear constraints using binary variables combined with power upper limits, ensuring the overall model can be stably solved using a standard mixed-integer linear programming solver. In the linearization of the absolute value of power change, the change in purchased power between adjacent time periods is first decomposed into positive and negative parts. Then, the absolute value is replaced by the sum of two non-negative auxiliary variables, expressed as: ; Mutual exclusion constraint linearization is achieved using binary variables combined with a power upper limit, expressed as follows using a battery as an example: ; In the formula, M is taken as the upper limit of the rated power of the equipment; Let the battery charge / discharge state be a binary variable, when Charging is only allowed at this time. Discharge is only permitted at this time.

[0067] The solution process takes into account daily load forecast data, renewable energy output forecast data, and electricity price forecast data, and ultimately obtains a baseline dispatch plan covering the entire dispatch cycle. The baseline dispatch plan includes a baseline process scheduling scheme and a baseline energy supply scheme, which correspond to the process operation arrangements and the operating power arrangements of various energy equipment in each time period throughout the entire cycle, respectively.

[0068] The baseline scheduling plan is subjected to intraday rolling optimization at a preset step size, and process timing constraints are applied within the rolling time domain to output the scheduling execution instruction for the current time period. The intraday rolling optimization is performed at a preset time step size. In this embodiment, a 1-hour rolling step size is used, and each re-optimization is completed within a preset prediction window. Only the scheduling execution instruction for the current time period is output, and the remaining time period scheduling plan obtained by the intraday rolling optimization is updated with the next rolling cycle.

[0069] Intraday rolling optimization is based on real-time state deviation and re-optimization, employing a rolling temporal optimization framework with a 1-hour rolling step and a prediction window length of 12 consecutive time periods. The prediction window length must satisfy the following condition: the window temporal domain must at least cover the shortest total processing time of the current process chain to be executed, ensuring that the window can fully accommodate all process timing constraints of unexecuted processes and avoiding constraint failure due to rolling truncation. In this embodiment, the total processing time of a single batch of withering, shaking, fixing, rolling, and drying is less than 12 hours. Therefore, a fixed 12-time-period window can fully encompass the timing constraints of all unexecuted processes within the entire scheduling cycle, ensuring model solvability and consistency with process rules. For scenarios with longer process chains, two adaptation methods can be adopted: one is to dynamically expand the window length according to the shortest total time of the current remaining processes; the other is to retain a fixed window length and embed the process states determined before the window start time, the remaining time of completed processes, and the running processes as hard boundary conditions into the model within the window, ensuring constraint continuity through boundary propagation. At the start of each rolling cycle, real-time status deviations on the energy side and the process side are collected and quantitatively updated. Energy side deviations include the deviation between the actual output of photovoltaic power generation and the day-ahead forecast, the deviation between the real-time grid price and the day-ahead price, and the deviation between the plant's basic electric and heat load and the forecast. The above deviation data are synchronously substituted into and updated in the corresponding input parameter sequence within the current forecast window. Process side deviations include the deviation between the actual start / end time of each process and the baseline scheduling plan, and the deviation between the actual energy consumption power of the running process and the rated parameters. For completed processes, their operating status and energy consumption data are locked as fixed quantities and will not participate in subsequent optimization adjustments.

[0070] During the current rolling cycle re-optimization process, the baseline process scheduling scheme and baseline energy supply scheme for subsequent periods within the forecast window are simultaneously revised: In terms of process scheduling, the start-up time of processes that have not yet started is allowed to be adjusted, provided all process timing constraints are met; for processes currently in operation, their continuous operation status is forcibly locked, and the remaining runtime constraints are updated based on the executed duration to ensure that the total runtime of the processes meets process requirements. In terms of energy supply, based on the updated load and output forecasts, the timing allocation of grid power purchase and sale, battery energy storage charging and discharging power, electric heating input power, biomass heating power, and thermal storage charging and discharging power is re-optimized. The adjustment process strictly ensures that the scheme always meets all process execution rules, including single-start of processes, continuous operation, and sequential connection of adjacent processes. Simultaneously, the timing shifting capability of energy storage mitigates operational fluctuations caused by deviations in renewable energy output and load forecasts, reducing the discrepancy between the scheduling plan and actual execution, and improving the practical feasibility of the scheduling scheme.

[0071] During the rolling optimization process, constraint control is implemented in time-domain segments: the process operation status and energy equipment status of the executed time period are locked with actual operating values ​​and no longer participate in optimization; the decision variables within the current prediction window fully retain all process timing constraints and energy system equipment operation constraints within the corresponding time period, ensuring that the optimization results within the window strictly conform to process rules and equipment boundaries; for time periods outside the window that have not yet entered optimization, the results of the day-ahead baseline scheduling plan are temporarily used as a reference, and will be included in the optimization solution when rolling to the corresponding time domain. This forms a consistent constraint closed loop of day-ahead global baseline guidance, strict constraint solution within the window, and rolling iterative synchronous correction: the optimization results output by each rolling cycle synchronously update the temporary baseline scheme for subsequent time periods, and the process rules and equipment constraints of the entire cycle are implemented step by step through segmented rolling, which avoids the process logic break caused by rolling time domain truncation and ensures that the intraday correction is consistent with the optimization direction of the day-ahead baseline. After each rolling cycle is completed, the actual operating data of the current time period is used as the initial state input for the next cycle, and the scheduling scheme is dynamically iterated and updated throughout the entire scheduling cycle through real-time status feedback.

[0072] Specifically, the intraday rolling optimization is performed sequentially: Baseline transfer and rolling initialization: The baseline scheduling plan obtained from the global solution of the previous day is used as the initial baseline and initial feasible solution for intraday rolling optimization. The full-cycle baseline scheme is distributed at the start of the scheduling day. Rolling optimization calculations are triggered at the trigger time of each rolling cycle, such as once per hour. Rolling window update and initial state assignment: Starting from the current time, 12 consecutive time periods are extracted to form the current prediction window. Two types of initial state assignments are completed simultaneously: on the energy side, the actual state of charge of battery storage and the actual heat storage state of thermal storage devices at the end of the previous rolling cycle are used as the initial state variables for this cycle's optimization; on the process side, the actual execution progress of each process is statistically analyzed, completed processes are marked as terminated and their running variables are locked, running processes are recorded for their running time and forced to maintain continuous operation, and processes that have not yet started retain the original allowed start time window constraints. Local rolling MILP model construction: Based on the updated prediction parameters and initial states, a local mixed integer linear programming model within the current prediction window is constructed. The model's decision variables are consistent with the day-ahead global model, including process scheduling variables, electrical power supply variables, and thermal power supply variables for each time period within the window. The model's constraints fully retain process timing constraints, electrical power balance constraints, thermal power balance constraints, electrothermal conversion constraints, and equipment operating boundary constraints, and adapt and modify the constraint boundaries of executed and running processes based on the initial state. The model's objective function uses the same weighted multi-objective function and weight coefficients as the day-ahead global model, ensuring that the intraday correction is consistent with the optimization direction of the day-ahead baseline. For model solving and optimization sequence correction, a standard mixed-integer linear programming solver is used to solve the local rolling MILP model, obtaining the process scheduling and energy supply optimization sequence within the current prediction window. This optimization sequence replaces the corresponding time period content in the original baseline scheduling plan, forming the corrected temporary baseline scheme. If an anomaly occurs in the local model where no feasible solution is found, a degraded solution strategy is automatically triggered: firstly, the upper and lower boundaries of the allowable start-up time window for the process are relaxed; secondly, the weight of the process completion time term in the multi-objective function is adjusted. Feasible solutions are obtained while strictly preserving constraints on equipment safe operation and the balance between electricity and heat supply and demand, thus avoiding scheduling interruptions. For the first time period's instruction output and execution, only the scheduling result of the first time period in the optimization sequence is retained as the scheduling execution instruction for the current time period. This instruction is sent to the energy management system and production control system for execution. Specifically, this includes power-side operation instructions such as grid-purchased power / grid-connected power, battery energy storage charging and discharging power, photovoltaic power generation local consumption power, and electric heating input power; heat-side operation instructions such as biomass heating power and thermal storage charging and discharging power; and process start / stop status instructions for the corresponding time period.Rolling iteration and status feedback: After the current time period is completed, the actual operating data for that time period is collected, including actual energy consumption, actual equipment output, and actual process execution status. The deviation from the optimization instructions is calculated, and this actual operating status is used as the initial state for the next rolling cycle. The rolling window is then updated to enter the next cycle of optimization, and this process is repeated until the optimization execution of the entire scheduling cycle is completed. Daily results summary: After the scheduling cycle ends, the actual operating data for the entire day is summarized, and core operating indicators are statistically analyzed and output, including total operating cost, total carbon emissions, net load peak, power purchase fluctuation rate, and total process completion time, for use in evaluating scheduling effectiveness and iterative calibration of subsequent optimization parameters.

[0073] This embodiment proposes a two-layer collaborative scheduling method that combines day-ahead global optimization with intraday rolling correction. By designing a consistent constraint solution mechanism between the upper and lower layers, the day-ahead plan is corrected in a closed loop based on the system state deviation during the intraday rolling process, which effectively improves the scheduling execution stability under the condition of renewable energy output and load forecast deviation.

[0074] Finally, it should be noted that the above content is only used to illustrate the technical solution of this application, and is not intended to limit the scope of protection of this application. Simple modifications or equivalent substitutions made by those skilled in the art to the technical solution of this application shall not depart from the substance and scope of the technical solution of this application.

Claims

1. A method for scheduling a tea-integrated distributed energy system, characterized in that, include: For the tea-making process with combined electric and thermal loads, a series production process chain is established, and an electric power demand model and a thermal power demand model are established for each production process. The process timing constraints of the series production process chain are set accordingly. Based on the electric power demand model and the thermal power demand model, a mixed-integer linear programming optimization model is constructed; the decision variables of the mixed-integer linear programming optimization model include process scheduling variables, electric power supply variables, and thermal power supply variables. The mixed-integer linear programming optimization model is solved globally one day in advance to obtain the baseline scheduling plan; The baseline scheduling plan is subjected to intraday rolling optimization at a preset step size, and the process timing constraints are applied in the rolling time domain to output the scheduling execution instruction for the current time period.

2. The tea-energy coordinated distributed energy system scheduling method according to claim 1, characterized in that, The process timing constraints include single-start constraints, continuous operation constraints, sequential connection constraints between adjacent processes, and allowed start time window constraints for processes.

3. The tea-energy coordinated distributed energy system scheduling method according to claim 1, characterized in that, In establishing the electrical power demand model and thermal power demand model for each production process, the operating status of each process is used as the input for model calculation; The process scheduling variables are used to obtain the running status of each process through time-series mapping.

4. The tea-energy coordinated distributed energy system scheduling method according to claim 3, characterized in that, The load-side input of the mixed-integer linear programming optimization model includes the total electrical load and total heat load of the tea-making process's serial production chain; The total electrical load is obtained by summing up the electrical power requirements of each process. The total heat load is obtained by summing up the heat power requirements of each process.

5. The tea-energy coordinated distributed energy system scheduling method according to claim 1, characterized in that, The constraints of the mixed-integer linear programming optimization model also include electrical power balance constraints, thermal power balance constraints, electrothermal conversion constraints, and equipment operation boundary constraints.

6. The tea-energy coordinated distributed energy system scheduling method according to claim 5, characterized in that, The mixed-integer linear programming optimization model uses a preset weighted multi-objective function as the objective function. The optimization objectives of the weighted multi-objective function include one or a combination of operating cost, carbon emission, net load peak, power purchase fluctuation, and process completion time.

7. The tea-energy coordinated distributed energy system scheduling method according to claim 1, characterized in that, The power supply variables on the electric side include grid power variables, battery energy storage charging and discharging power variables, renewable energy consumption power variables, and electric heating input power variables.

8. The tea-energy coordinated distributed energy system scheduling method according to claim 7, characterized in that, The heat-side energy supply variables include biomass heating power variables, thermal storage charging and releasing power variables, and electric heating output power variables; Based on the electrothermal conversion efficiency, the electric heating output power is calculated according to the electric heating input power of the series production process chain in the tea-making process.

9. The tea-energy coordinated distributed energy system scheduling method according to claim 1, characterized in that, The benchmark scheduling plan includes a benchmark process scheduling scheme and a benchmark energy supply scheme; The intraday rolling optimization is performed according to a preset time step, and each re-optimization is completed within a preset prediction window to output the scheduling execution instruction for the current time period. The scheduling plan obtained by the intraday rolling optimization is updated with the next rolling cycle.

10. The tea-energy coordinated distributed energy system scheduling method according to claim 9, characterized in that, The intraday rolling optimization is based on real-time state deviations and re-optimization is performed. In the intraday rolling optimization, the baseline process scheduling scheme and baseline energy supply scheme for subsequent periods are simultaneously modified within or after the current rolling cycle.

Images