Intelligent energy scheduling method and system for food production

Abstract

The invention discloses an intelligent energy scheduling method and system for food production, and relates to the technical field of energy scheduling, and the method comprises the steps: collecting an original multi-source data set, carrying out the preprocessing, and generating a standardized thermal data set; identifying a process hot water demand, a secondary sterilization hot water demand, a cooling load demand and a waste heat resource from the standardized thermotechnical data set to form a heat supply and demand object set; constructing a virtual energy network based on the predicted supply and demand sequence, performing supply and demand mapping and constraint association on the virtual energy network, and establishing an optimal scheduling model; and controlling a heat pump, a circulating pump, a valve and a boiler to operate according to the optimal heat scheduling strategy, collecting operation feedback data and a process standard reaching result, performing rolling correction on the optimal scheduling model based on the operation feedback data and the process standard reaching result, and outputting a correction scheduling instruction. According to the method, collaborative optimization solution is carried out again, and the corrected scheduling instruction is output, so that the possibility of misalignment of the scheduling result in the long-term operation process is reduced.

Description

Technical Field

[0001] This invention relates to the field of energy dispatching technology, and in particular to an intelligent energy dispatching method and system for food production. Background Technology

[0002] With the development of continuous, automated, and green manufacturing in the food industry, energy utilization in food production has gradually evolved from single heating or cooling to a coordinated heating and cooling energy supply approach covering the entire production process. This is particularly true in the production of fruit juices, dairy products, condiments, and ready-to-eat foods, where processes such as raw material preheating, process heating, secondary sterilization, cooling and shaping, and workshop environmental regulation often coexist. These processes involve significant simultaneous heat and cooling load demands. Furthermore, with the development of sensor networks, energy metering devices, and industrial data platforms, production planning data, temperature data, flow data, pressure data, power data, and environmental monitoring data are being collected and analyzed in a unified manner, providing a data foundation for predicting energy demand, identifying waste heat, and optimizing energy scheduling in the food production process.

[0003] However, existing technologies still have the following shortcomings: existing energy dispatching technologies for food production mainly rely on boiler steam heat exchange or single heat pump heating, resulting in extensive heat source configuration and low degree of cold and heat coupling in the food production process. Although existing solutions can partially recover waste heat from cooling water and hot air in the factory, they usually remain at the level of fixed path recovery, making it difficult to achieve the cascade utilization of waste heat of different grades and dynamic matching between heat pumps, boilers, and heat storage units. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent energy scheduling method for food production to solve the problems of extensive heat source configuration and low thermal coupling in the food production process.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent energy scheduling method for food production, comprising,

[0008] Collect raw multi-source datasets and preprocess them to generate standardized thermal datasets. Identify process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal datasets to form a set of heat supply and demand objects.

[0009] Based on the heat supply and demand object set and standardized thermal data set, the heat load curve, cold load curve, waste heat output and waste heat grade are predicted within the future time window to obtain the predicted supply and demand sequence.

[0010] A virtual energy network is constructed based on predicted supply and demand sequences. Supply and demand mapping and constraint correlation are performed on the virtual energy network, and an optimal scheduling model is established.

[0011] Pre-peak thermal storage pre-scheduling is performed on the thermal storage capacity in the virtual energy network to generate pre-peak thermal storage results. Different grades of waste heat in the predicted supply and demand sequence, as well as the heat pump heating capacity, boiler supplementary heating capacity and pre-peak thermal storage results in the virtual energy network, are input into the optimization scheduling model for collaborative optimization and solution, and the optimal heat scheduling strategy is output.

[0012] The operation of heat pumps, circulating pumps, valves and boilers is controlled according to the optimal heat dispatch strategy. Operational feedback data and process compliance results are collected. The optimized dispatch model is rolled over based on the operational feedback data and process compliance results, and the corrected dispatch instructions are output.

[0013] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for generating the standardized thermal dataset are as follows:

[0014] Obtain the raw multi-source dataset, which includes production planning data, energy metering data, supply and return water temperature data, flow data, pressure data, power data, and environmental monitoring data;

[0015] Unify the timestamps of the original multi-source datasets to form a unified timeline data;

[0016] The data on the unified time axis are filled with missing items, abnormal fluctuation items are removed, and the units are unified to obtain standardized monitoring data.

[0017] Standardized monitoring data are associated and labeled according to the production line number, process section number, equipment number, and scheduling time window in the production process to generate a standardized thermal dataset.

[0018] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for forming a heat supply and demand object set are as follows:

[0019] Extract the process hot water demand, secondary sterilization hot water demand, and cooling load demand in the food production process from the standardized thermal dataset to form a demand object set;

[0020] Waste heat resources are obtained by extracting waste heat from cooling water and waste heat from hot air in the plant from standardized thermal datasets.

[0021] The demand object set and waste heat resources are identified and aggregated to generate a heat supply and demand object set.

[0022] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for obtaining the predicted supply and demand sequence are as follows:

[0023] The heat supply and demand object set is segmented according to the scheduling time window to generate time window supply and demand segments;

[0024] By using time windows to filter historical correlation segments under the same batch, process section, and environmental conditions from the standardized thermal data set;

[0025] The supply and demand segments within a time window are matched with historical related segments to generate target related segments;

[0026] Based on the heat and cold load change trends of the target associated segments, the process hot water demand, secondary sterilization hot water demand, and cooling load demand are extrapolated to generate heat load curves and cold load curves.

[0027] Based on the cooling water temperature difference, ambient air temperature difference, and heat pump operating status in the standardized thermal data set, waste heat assessment is performed on the heat load curve and cooling load curve to determine the waste heat output and waste heat grade.

[0028] By combining the heat load curve, cooling load curve, waste heat output, and waste heat grade, a predicted supply and demand sequence is generated.

[0029] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for constructing a virtual energy network based on predicted supply and demand sequences are as follows:

[0030] Based on the process hot water demand, secondary sterilization hot water demand, and cooling load demand, process hot water node, secondary sterilization node, and cooling node are established respectively.

[0031] Waste heat from cooling water and waste heat from hot air in the plant are classified and labeled by waste heat output and waste heat grade to generate waste heat nodes.

[0032] Based on the relationships between preheating, temperature raising, deep heating, heat replenishment and heat storage in the food production process, air source heat pump nodes, high temperature heat pump nodes, boiler nodes and heat storage nodes are established respectively.

[0033] By linking process hot water nodes, secondary sterilization nodes, cooling nodes, waste heat nodes, air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes, a virtual energy network is formed.

[0034] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for establishing the optimized scheduling model are as follows:

[0035] Read the node types and node attributes in the virtual energy network, and map process hot water nodes, secondary sterilization nodes, and cooling nodes to demand-side nodes based on node types and node attributes, and map waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and thermal storage nodes to supply-side nodes. Pair demand-side nodes and supply-side nodes according to scheduling time windows to generate supply and demand mapping relationships.

[0036] The heating sequence, heat replenishment sequence, and heat storage sequence in the supply and demand mapping relationship are configured with time sequence association to generate the supply and demand relationship;

[0037] Based on the supply and demand relationship, configure process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints, and generate constraint relationship results;

[0038] By integrating the supply and demand mapping relationship, the supply and demand correlation relationship, and the constraint correlation results, the heat scheduling object, scheduling order, scheduling boundary and scheduling objective are determined, and an optimized scheduling model is formed.

[0039] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for generating pre-peak thermal storage results are as follows:

[0040] Identify peak time windows based on the heat load curve in the predicted supply and demand sequence;

[0041] Extract waste heat output from the scheduling time window before the peak time window, and conduct joint analysis of heat pump heating capacity, boiler supplementary heating capacity, heat storage capacity and waste heat output in the virtual energy network to identify the pre-storage time window.

[0042] Peak demand during peak time windows and available heat during pre-storage time windows are shaving and valley filling to generate heat storage.

[0043] The heat storage capacity is combined with the heat storage capacity in the pre-storage time window to form the pre-peak heat storage result.

[0044] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for outputting the optimal heat scheduling strategy are as follows:

[0045] The waste heat of different grades in the predicted supply and demand sequence is classified to generate waste heat classification results. The waste heat classification results are sequentially matched and heat is allocated through the heat pump heating capacity, boiler supplementary heating capacity and pre-peak heat storage results in the virtual energy network to generate initial scheduling results.

[0046] The initial scheduling results are verified based on the constraint association results, and the optimal heat scheduling strategy is generated.

[0047] As a preferred embodiment of the intelligent energy scheduling method for food production described in this invention, the specific steps for collecting operational feedback data and process compliance results are as follows:

[0048] The optimal heat dispatch strategy is converted into heat pump start / stop commands, circulating pump speed adjustment commands, valve switching commands, and boiler intervention commands, which are then sent to the heat pump, circulating pump, valve, and boiler for operation.

[0049] The system collects data on temperature, flow rate, pressure, power, and process compliance during the operation of heat pumps, circulating pumps, valves, and boilers, and obtains operational feedback data and process compliance results.

[0050] Secondly, the present invention provides an intelligent energy dispatching system for food production, comprising,

[0051] The heat supply and demand module is used to collect raw multi-source datasets, preprocess them, generate standardized thermal datasets, identify process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal datasets, and form a heat supply and demand object set.

[0052] The supply and demand sequence module is used to predict the heat load curve, cooling load curve, waste heat output and waste heat grade within a future time window based on the heat supply and demand object set and the standardized thermal data set, so as to obtain the predicted supply and demand sequence.

[0053] The scheduling model module is used to construct a virtual energy network based on the predicted supply and demand sequence, perform supply and demand mapping and constraint association on the virtual energy network, and establish an optimized scheduling model.

[0054] The pre-peak thermal storage module is used to pre-schedule the thermal storage capacity in the virtual energy network, generate pre-peak thermal storage results, and input the waste heat of different grades in the predicted supply and demand sequence, the heat pump heating capacity and boiler supplementary heating capacity in the virtual energy network, and the pre-peak thermal storage results into the optimization scheduling model for collaborative optimization and solution, and output the optimal heat scheduling strategy.

[0055] The correction scheduling module is used to control the operation of heat pumps, circulating pumps, valves and boilers according to the optimal heat scheduling strategy, and to collect operation feedback data and process compliance results. Based on the operation feedback data and process compliance results, the module performs rolling corrections to the optimized scheduling model and outputs correction scheduling instructions.

[0056] The beneficial effects of this invention are as follows: By performing timestamp unification, missing item completion, abnormal fluctuation item removal, dimension unification, and correlation labeling on the original multi-source dataset, a standardized thermal dataset is generated, improving the reliability of prediction and scheduling results. By classifying and labeling the waste heat of cooling water and the waste heat of hot air in the plant and forming a virtual energy network, the degree of relationship confusion during multi-device collaborative scheduling is reduced. By configuring the time-series correlation of heating sequence, heat replenishment sequence, and heat storage sequence, an optimized scheduling model is formed, improving the orderliness of supply and demand matching and the constraint integrity of scheduling solutions. By performing collaborative optimization again, corrected scheduling instructions are output, reducing the possibility of inaccurate scheduling results during long-term operation. Attached Figure Description

[0057] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a flowchart of an energy intelligent scheduling method for food production.

[0059] Figure 2 This is a schematic diagram of an intelligent energy dispatching system used in food production.

[0060] Figure 3 A flowchart for establishing an optimized scheduling model.

[0061] Figure 4 The flowchart for outputting the corrected scheduling instructions. Detailed Implementation

[0062] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0063] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0064] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0065] like Figure 1 As shown, this is an embodiment of the present invention, which provides an intelligent energy scheduling method for food production, including the following steps:

[0066] S1: Collect the original multi-source dataset and preprocess it to generate a standardized thermal dataset. Identify the process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal dataset to form a heat supply and demand object set.

[0067] S1.1: Obtain the original multi-source dataset, unify the timestamps of the original multi-source dataset to form a unified timeline data; fill in missing items, remove abnormal fluctuation items and unify the units of measurement of the unified timeline data to obtain standardized monitoring data.

[0068] Production planning data is obtained from the production planning management terminal in the food production process; energy metering data is obtained from energy metering devices; supply and return water temperature data is obtained from supply and return water monitoring points; flow rate data is obtained from flow rate monitoring points; pressure data is obtained from pressure monitoring points; power data is obtained from power monitoring points; and environmental monitoring data is obtained from plant environmental monitoring points. These data are then aggregated to form a raw multi-source dataset. The acquisition period of the raw multi-source dataset is calculated, and the smallest time unit in the acquisition period is used as the alignment benchmark. The earliest acquisition time in the raw multi-source dataset is used as the start time, and the latest acquisition time is used as the end time. Starting from the start time, time points are generated incrementally according to the alignment benchmark until the end time, resulting in a continuous sequence of time points arranged in chronological order. The acquisition time of each data record in the raw multi-source dataset is then read. The collected data time is uniformly converted to the format of year-month-day-hour-minute-second. The uniformly converted collected time is compared with the time points in the continuous time series one by one. When the collected time of a data point among the production planning data, energy metering data, supply and return water temperature data, flow data, pressure data, power data, and environmental monitoring data is consistent with a time point in the continuous time series, this data point is written to the corresponding position in the continuous time series. When the collected time of a data point is between two time points in the continuous time series, this data point is written to the corresponding position of the previous time point. The production planning data, energy metering data, supply and return water temperature data, flow data, pressure data, power data, and environmental monitoring data written to the corresponding positions of the same time point are arranged in the order of data source to form a unified time axis data.

[0069] The missing positions in the unified timeline data are identified, and the missing items are filled in using the adjacent valid data before and after the missing positions to form the completed unified timeline data. The data changes at each time point in the completed unified timeline data are read, and the direction and continuity of data changes between adjacent time points are compared. Data that disrupts the continuous trend is identified as abnormal fluctuations, which are removed from the completed unified timeline data and replaced with adjacent valid data before and after the abnormal fluctuations to form purified unified timeline data. The temperature data, flow data, pressure data, power data, and energy metering data in the purified unified timeline data are converted into unified measurement units to obtain standardized monitoring data.

[0070] It should be noted that the production planning data, energy metering data, supply and return water temperature data, flow data, pressure data, power data, and environmental monitoring data in the original multi-source dataset belong to multi-source heterogeneous thermal data in the food production process. After time unification, anomaly cleaning, dimension unification, and correlation labeling, the multi-source heterogeneous thermal data is transformed into a standardized thermal dataset. The standardized thermal dataset serves as the data analysis basis for subsequent heat supply and demand identification, historical correlation segment screening, heat load prediction, cold load prediction, and waste heat assessment.

[0071] S1.2: Correlate and label the standardized monitoring data according to the production line number, process section number, equipment number and scheduling time window in the production process to generate a standardized thermal dataset.

[0072] Read each data record from the standardized monitoring data, including its acquisition location, acquisition time, and data type. Also read the equipment installation location relationships, production line equipment configuration relationships, and scheduling time windows in the production process, as well as the process start and end times from the production plan data. Based on the acquisition location and equipment installation location relationships, determine the equipment number corresponding to each data record. Based on the equipment number and production line equipment configuration relationships, determine the production line number corresponding to each data record. Based on the acquisition time's interval between the process start and end times, determine the process segment number for each data record. Based on the acquisition time's position within the scheduling time window (e.g., 1 minute or 5 minutes), determine the scheduling time window to which each data record belongs. Write the equipment number, production line number, process segment number, and scheduling time window into each data record to generate a standardized thermal dataset.

[0073] S1.3: Extract the process hot water demand, secondary sterilization hot water demand, and cooling load demand from the standardized thermal data set to form a demand object set; extract the waste heat of cooling water and waste heat of hot air in the plant from the standardized thermal data set to obtain waste heat resources.

[0074] Standardized thermal data belonging to the heating process in food production is extracted from the production line number and process section number. The supply water temperature, return water temperature, and process hot water flow rate of the heating process section are read. For example, the supply water temperature is 60℃~80℃, the return water temperature is 40℃~70℃, and the process hot water flow rate is 2m³ / h~10m³ / h, forming the process hot water demand. The heating process section can be any one of the following: the preliminary heating process for juice filling, the preheating process for dairy products, or the heat processing process for condiments. Standardized thermal data belonging to the secondary sterilization process is also extracted from the production line number and process section number. The hot water temperature, circulation flow rate, and energy metering data of the secondary sterilization process section are read. For example, the hot water temperature is 70℃~75℃, and the circulation flow rate is 3m³ / h~8m³ / h, forming the secondary sterilization hot water demand. Standardized thermal data belonging to the cooling process are filtered out from equipment numbers and scheduling time windows. Supply and return water temperatures, circulation flow rates, and environmental monitoring data during the cooling process are read. For example, if the supply and return water temperatures are 10℃~35℃ and the circulation flow rates are 5m³ / h~20m³ / h, cooling load demand is formed. The process hot water demand, secondary sterilization hot water demand, and cooling load demand in the food production process are collected according to production line number, process section number, and scheduling time window to form a demand object set. Cooling water circulation data is filtered out from equipment numbers, and the supply and return water temperatures and flow rates in the cooling water circulation data are read to form cooling water waste heat. Plant environmental data is filtered out from environmental monitoring data, and the air temperature in the plant environmental data is read to form plant hot air waste heat. Cooling water waste heat and plant hot air waste heat are organized according to scheduling time windows to obtain waste heat resources.

[0075] S1.4: Identify and aggregate the demand object set and waste heat resources to generate a heat supply and demand object set.

[0076] The process hot water demand in the demand object set is identified as the process hot water demand type, the secondary sterilization hot water demand is identified as the secondary sterilization hot water demand type, the cooling load demand is identified as the cooling load demand type, the cooling water waste heat in the waste heat resources is identified as the cooling water waste heat type, and the plant hot air waste heat is identified as the plant hot air waste heat type. The cooling water waste heat and plant hot air waste heat are identified according to the scheduling time window. The process hot water demand, secondary sterilization hot water demand, cooling load demand, cooling water waste heat and plant hot air waste heat that have completed the type identification are aggregated according to the production line number, process section number and scheduling time window to generate a heat supply and demand object set.

[0077] S2: Based on the heat supply and demand object set and standardized thermal data set, predict the heat load curve, cold load curve, waste heat output and waste heat grade within the future time window to obtain the predicted supply and demand sequence.

[0078] S2.1: Divide the heat supply and demand object set into segments according to the scheduling time window to generate time window supply and demand segments; filter historical related segments of the same batch, the same process section, and the same environmental conditions from the standardized thermal data set through the time window supply and demand segments.

[0079] The process hot water demand, secondary sterilization hot water demand, cooling load demand, cooling water waste heat, and plant hot air waste heat with the same production line number, process section number, and scheduling time window are grouped into the same time period to generate a time window supply and demand segment. The batch characteristics, process section number, and environmental monitoring characteristics in the time window supply and demand segment are compared item by item with the batch number, process section number, and environmental monitoring data in the standardized thermal data set. Data in the standardized thermal data set that are consistent with the batch characteristics, process section number, and environmental temperature characteristics are selected to generate historical correlation segments. The length of the historical correlation segment is consistent with the length of the time window supply and demand segment. The scheduling time window is set to 1 minute, 5 minutes, or 10 minutes according to the production data refresh frequency and heat adjustment frequency.

[0080] S2.2: Perform similarity matching between the supply and demand segments of the time window and historical related segments to generate target related segments.

[0081] Numerical information on process hot water demand, secondary sterilization hot water demand, cooling load demand, cooling water waste heat, and plant hot air waste heat is extracted from the supply and demand segments within the time window. Similarly, numerical information on these same parameters is extracted from historical correlation segments. The differences between all numerical information in each time window supply and demand segment and all numerical information in each historical correlation segment are calculated to obtain the differences in process hot water demand, secondary sterilization hot water demand, cooling load demand, cooling water waste heat, and plant hot air waste heat. These differences are then aggregated according to the same historical segment to form a degree of difference, completing a multidimensional correlation analysis. The historical correlation segment with the smallest degree of difference is identified as the target correlation segment. When abnormal operating conditions caused by shutdowns, start-stop switching, sudden changes in supply and return water, or environmental changes exist in the historical correlation segments, the historical correlation segments containing abnormal operating conditions are removed, and the degree of difference is calculated again.

[0082] S2.3: Based on the heat and cold change trends of the target associated segments, extrapolate the trends of process hot water demand, secondary sterilization hot water demand, and cooling load demand to generate heat load curves and cold load curves.

[0083] Extract the heat and cooling load change trends between adjacent scheduling time windows from the target associated segments and arrange them in the order of the scheduling time windows. Take the process hot water demand and secondary sterilization hot water demand in the current scheduling time window as the initial heat value and the cooling load demand in the current scheduling time window as the initial cooling load value. Accumulate the heat change trends to the initial heat values ​​in the order of the scheduling time windows to obtain the predicted values ​​of process hot water demand and secondary sterilization hot water demand in the scheduling time window. Accumulate the cooling load change trends to the initial cooling load values ​​in the order of the cooling load to obtain the predicted values ​​of cooling load demand in the scheduling time window. Connect the predicted values ​​of process hot water demand and secondary sterilization hot water demand in the order of the scheduling time windows to generate a heat load curve. Connect the predicted values ​​of cooling load demand in the order of the scheduling time windows to generate a cooling load curve. When the production plan data is adjusted by batch, process segment, or time, the corresponding historical associated segments are re-filtered, and the heat load curve and cooling load curve are updated by the re-filtered historical associated segments.

[0084] S2.4: Based on the cooling water temperature difference, ambient air temperature difference, and heat pump operating status in the standardized thermal data set, perform waste heat assessment on the heat load curve and cold load curve to determine the waste heat output and waste heat grade; combine the heat load curve, cold load curve, waste heat output, and waste heat grade to generate a predicted supply and demand sequence.

[0085] Cooling water temperature difference, ambient air temperature difference, flow rate, and heat pump operating status are extracted from the standardized thermal dataset and arranged in the order of scheduling time windows. The cooling water temperature difference, ambient air temperature difference, flow rate, and heat pump operating status are matched one by one with the cooling load curve under the same scheduling time window. Scheduling time windows with cooling water temperature difference greater than zero, ambient air temperature difference greater than zero, flow rate in circulation, and heat pump operating status in operation are screened out from the cooling load curve. The recyclable cooling load portion in the scheduling time window is taken as waste heat output. The cooling water temperature difference, ambient air temperature difference, and flow rate in the scheduling time window that forms waste heat output are sorted by value, and the scheduling time window with heat pump operating status in operation is placed first to form waste heat grade. The heat load curve, cooling load curve, waste heat output, and waste heat grade in all scheduling time windows are collected in chronological order to generate a predicted supply and demand sequence. At the same time, the window length of the trend extrapolation is consistent with the future time window length of the predicted supply and demand sequence.

[0086] like Figure 3 As shown, S3: Construct a virtual energy network based on the predicted supply and demand sequence, perform supply and demand mapping and constraint association on the virtual energy network, and establish an optimal scheduling model.

[0087] S3.1: Based on the process hot water demand, secondary sterilization hot water demand, and cooling load demand, establish process hot water nodes, secondary sterilization nodes, and cooling nodes respectively; classify and label the waste heat of cooling water and waste heat of plant hot air by waste heat output and waste heat grade to generate waste heat nodes.

[0088] The production line number, process section number, and scheduling time window are extracted from the process hot water demand, secondary sterilization hot water demand, and cooling load demand, respectively. The process hot water demand, its corresponding production line number, process section number, and scheduling time window are combined to form a process hot water node. The secondary sterilization hot water demand, its corresponding production line number, process section number, and scheduling time window are combined to form a secondary sterilization node. The cooling load demand, its corresponding production line number, process section number, and scheduling time window are combined to form a cooling node. Waste heat output and waste heat grade are extracted from the predicted supply and demand sequence. Cooling water waste heat and plant hot air waste heat are extracted from waste heat resources. The waste heat grade and waste heat output corresponding to the cooling water waste heat are written into the identification field of the cooling water waste heat to form a cooling water waste heat identification result. The waste heat grade and waste heat output corresponding to the plant hot air waste heat are written into the identification field of the plant hot air waste heat to form a plant hot air waste heat identification result. The cooling water waste heat identification result and the plant hot air waste heat identification result are summarized to generate a waste heat node.

[0089] S3.2: Based on the relationships between preheating, temperature raising, deep heating, heat replenishment and heat storage in the food production process, establish air source heat pump nodes, high temperature heat pump nodes, boiler nodes and heat storage nodes respectively.

[0090] The corresponding action locations and durations of each heating link are extracted from the preheating, temperature rise, deep heating, heat replenishment, and heat storage relationships in the food production process. The action locations, durations, and heat pump heating capacities corresponding to the preheating and temperature rise relationships are combined into a single node record to form an air source heat pump node. The action locations, durations, and heat pump heating capacities corresponding to the deep heating relationship are combined into a single node record to form a high-temperature heat pump node. The action locations, durations, and boiler heat replenishment capacities corresponding to the heat replenishment relationship are combined into a single node record to form a boiler node. The action locations, durations, and heat storage capacities corresponding to the heat storage relationship are combined into a single node record to form a heat storage node.

[0091] S3.3: Connect process hot water nodes, secondary sterilization nodes, cooling nodes, waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and thermal storage nodes to form a virtual energy network.

[0092] Read the production line number, process section number, scheduling time window, action location, and action period from the process hot water node, secondary sterilization node, cooling node, waste heat node, air source heat pump node, high temperature heat pump node, boiler node, and thermal storage node. Pair nodes with connected action locations and consecutive action periods to form node connection pairs. Establish connections between the waste heat node and the air source heat pump node, high temperature heat pump node, and thermal storage node respectively. Connect the air source heat pump node, high temperature heat pump node, boiler node, and thermal storage node to the process hot water node, secondary sterilization node, and cooling node respectively. Connect the waste heat node to the air source heat pump node, high temperature heat pump node, and thermal storage node respectively. The connections formed by the nodes, as well as the connections formed by air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes with process hot water nodes, secondary sterilization nodes, and cooling nodes, are arranged in the order of scheduling time windows to form a node connection chain. The output end of the previous node in the node connection chain is connected to the input end of the next node to form a network connection structure that includes heat input path, heat conversion path, heat replenishment path, and heat output path. The process hot water nodes, secondary sterilization nodes, cooling nodes, waste heat nodes, air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes in the network connection structure are brought together to form a virtual energy network.

[0093] S3.4: Read the node types and node attributes in the virtual energy network, and map process hot water nodes, secondary sterilization nodes, and cooling nodes to demand-side nodes through node types and node attributes, and map waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and thermal storage nodes to supply-side nodes. Pair demand-side nodes and supply-side nodes according to scheduling time windows to generate supply and demand mapping relationships.

[0094] Process hot water nodes, secondary sterilization nodes, and cooling nodes are screened from the virtual energy network. Demand time periods, demand intensity, and temperature requirements are extracted from the node attributes of these nodes. The node types of these nodes are combined with the demand time periods, demand intensity, and temperature requirements to form demand-side identifiers. Process hot water nodes, secondary sterilization nodes, and cooling nodes with demand-side identifiers are identified as demand-side nodes. Waste heat nodes, air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes are also screened from the virtual energy network. Demand time periods, demand intensity, and temperature requirements are extracted from the node attributes of these nodes. By taking the waste heat grade, heat pump heating capacity, boiler supplementary heating capacity, and heat storage capacity, the node types of waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and heat storage nodes are combined with the waste heat grade, heat pump heating capacity, boiler supplementary heating capacity, and heat storage capacity to form a supply-side identifier. Waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and heat storage nodes with supply-side identifiers are identified as supply-side nodes. Demand-side nodes and supply-side nodes are paired according to the scheduling time window sequence to generate a supply-demand mapping relationship. Among them, demand-side nodes include heat demand-side nodes and cooling load demand-side nodes. Process hot water nodes and secondary sterilization nodes belong to heat demand-side nodes, and cooling nodes belong to cooling load demand-side nodes.

[0095] S3.5: Configure the heating sequence, heat replenishment sequence, and heat storage sequence in the supply and demand mapping relationship in a time sequence to generate the supply and demand relationship.

[0096] The demand-side nodes and supply-side nodes in the supply-demand mapping relationship are arranged according to the order of scheduling time windows. Demand-side nodes and supply-side nodes within the same scheduling time window are grouped into the same time sequence unit. Within the same time sequence unit, waste heat nodes are preferentially matched with air source heat pump nodes and high-temperature heat pump nodes to determine the order in which waste heat enters the air source heat pump nodes and high-temperature heat pump nodes, forming a heating sequence. The demand intensity of the demand-side nodes within the same time sequence unit is compared with the heat pump heating capacity and heat storage capacity of the supply-side nodes. When the heat pump heating capacity and heat storage capacity can cover the demand intensity, the heat storage node is placed next to the air source heat pump node and high-temperature heat pump node. Subsequently, a heat storage sequence is formed. When the heat pump's heating capacity and heat storage capacity are lower than the demand intensity, the boiler node is placed after the air source heat pump node, high-temperature heat pump node, and heat storage node to form a supplementary heating sequence. The heating sequence, supplementary heating sequence, and heat storage sequence within the same time unit are unfolded according to the process hot water node, secondary sterilization node, and cooling node to which the demand-side node belongs. This ensures that the process hot water node, secondary sterilization node, and cooling node each correspond to a set of consecutive heating sequence, supplementary heating sequence, and heat storage sequence within their respective scheduling time windows. The consecutive heating sequence, supplementary heating sequence, and heat storage sequence within adjacent scheduling time windows are connected end to end to form a supply and demand relationship.

[0097] S3.6: Configure process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints based on the supply and demand relationship, and generate constraint correlation results; integrate the supply and demand mapping relationship, supply and demand relationship, and constraint correlation results to determine the heat scheduling object, scheduling order, scheduling boundary, and scheduling objective, and form an optimized scheduling model.

[0098] Read the scheduling time windows, node sequence, and node capacity information of process hot water nodes, secondary sterilization nodes, cooling nodes, waste heat nodes, air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes from the supply and demand relationship. Write the temperature requirements corresponding to process hot water nodes and secondary sterilization nodes into the supply and demand relationship to form process temperature constraints. Write the scheduling time window sequence corresponding to process hot water nodes, secondary sterilization nodes, and cooling nodes into the supply and demand relationship to form process timing constraints. Write the heat pump heating capacity, boiler supplementary heating capacity, and thermal storage capacity corresponding to air source heat pump nodes, high-temperature heat pump nodes, boiler nodes, and thermal storage nodes into the supply and demand relationship to form equipment capacity constraints. Include supply and return water temperature data... The supply and return water status corresponding to flow and pressure data is written into the supply and demand relationship to form supply and return water pressure constraints. The intervention position and intervention time of the boiler node in the heating sequence are written into the supply and demand relationship to form boiler standby constraints. The process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints are aggregated according to the scheduling time window to generate constraint association results. The demand-side nodes and supply-side nodes in the supply and demand mapping relationship, the heating sequence, heat replenishment sequence, and heat storage sequence in the supply and demand relationship, and the various constraints in the constraint association results are uniformly arranged according to the scheduling time window to determine the heat scheduling object, scheduling sequence, scheduling variables, state variables, scheduling boundary, and scheduling objective, forming an optimized scheduling model.

[0099] It should be noted that the optimization scheduling model includes scheduling objectives, scheduling variables, state variables, and constraints. The scheduling objectives include at least one of reducing primary energy consumption, increasing waste heat utilization, and meeting process compliance requirements. The scheduling variables include at least one of the following: heat pump start-up / shutdown status, heat pump heating capacity, boiler heat replenishment, thermal storage charging / discharging, circulating pump flow distribution, and valve switching status. The state variables include at least one of the following: thermal storage status, heat pump operating status, boiler operating status, and process stage status. The constraints include at least heat balance constraints, process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints. The optimization scheduling model constructs a scheduling problem based on the scheduling objectives, scheduling variables, state variables, and constraints, and solves it using mixed integer programming.

[0100] S4: Perform pre-peak heat storage pre-scheduling on the heat storage capacity in the virtual energy network, generate pre-peak heat storage results, input different grades of waste heat in the predicted supply and demand sequence, heat pump heating capacity, boiler supplementary heating capacity and pre-peak heat storage results in the virtual energy network into the optimization scheduling model for collaborative optimization solution, and output the optimal heat scheduling strategy.

[0101] S4.1: Identify peak time windows based on the heat load curve in the predicted supply and demand sequence; extract waste heat output from the scheduling time window before the peak time window; and conduct joint analysis of the heat pump heating capacity, boiler supplementary heating capacity, heat storage capacity and waste heat output in the virtual energy network to identify the pre-storage time window.

[0102] The heat load curve values ​​are arranged in chronological order according to the scheduling time windows. The heat load curve values ​​in adjacent scheduling time windows are compared item by item. The scheduling time window with a heat load curve value higher than the previous scheduling time window and higher than the next scheduling time window is marked as the peak heat consumption time window. The time interval in which consecutive peak heat consumption time windows appear is determined as the peak time window. The waste heat output is extracted from each scheduling time window before the peak time window and arranged in chronological order according to the scheduling time windows. The heat pump heating capacity, boiler supplementary heating capacity, and heat storage capacity are extracted from the virtual energy network and expanded according to the scheduling time windows. The scheduling time window with waste heat output greater than zero, heat pump heating capacity and boiler supplementary heating capacity with available margin, and heat storage capacity with remaining capacity is marked as the pre-heat storage time window.

[0103] S4.2: Peak load demand during peak time windows and available heat during pre-storage time windows are shaving and valley filling to generate heat storage; the heat storage is combined with the heat storage capacity in the pre-storage time window to form pre-peak heat storage results.

[0104] The heat load demand in the peak time window is arranged according to the order of the scheduling time windows, and the available heat in the pre-storage time window is arranged according to the order of the scheduling time windows. The available heat in the pre-storage time window is transferred to the peak time window one by one, so that the heat load demand in the peak time window is made up in advance by the available heat in the pre-storage time window. The heat transferred from the pre-storage time window to the peak time window is determined as the heat storage heat. The heat storage capacity in the pre-storage time window is read, and the heat storage heat is mapped one by one to the heat storage capacity in the pre-storage time window to form the pre-peak heat storage result.

[0105] S4.3: Classify the waste heat of different grades in the predicted supply and demand sequence to generate waste heat classification results. Through the heat pump heating capacity, boiler supplementary heating capacity and pre-peak heat storage results in the virtual energy network, perform sequential matching and heat allocation on the waste heat classification results to generate initial scheduling results.

[0106] The waste heat output and grade within each scheduling time window are extracted from the predicted supply and demand sequence. The waste heat grades are sorted from high to low and assigned a grade number according to the sorting order to form a waste heat grading result. The waste heat grading result, the heat pump heating capacity, boiler supplementary heating capacity, and pre-peak heat storage result in the virtual energy network are input into the optimization scheduling model. The optimization scheduling model determines the demand-side nodes, supply-side nodes, heating sequence, supplementary heating sequence, and heat storage sequence participating in the scheduling according to the supply and demand mapping relationship. According to the constraint association result, the temperature range, time sequence range, capacity range, and pressure range within each scheduling time window are restricted. The optimization scheduling model allocates the heat pump heating capacity, boiler supplementary heating capacity, and pre-peak heat storage result to the process hot water node and secondary sterilization node according to the grade number in the waste heat grading result to generate the initial scheduling result.

[0107] S4.4: Verify the initial scheduling results based on the constraint association results, and generate the optimal heat scheduling strategy.

[0108] The initial scheduling results, including the heating sequence, heat replenishment sequence, heat storage sequence, and heat allocation, are compared with the constraint association results, including process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints. When the heating sequence, heat replenishment sequence, heat storage sequence, and heat allocation results simultaneously satisfy the process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints, the initial scheduling results are determined as the optimal heat scheduling strategy. Otherwise, the heating sequence, heat replenishment sequence, heat storage sequence, or heat allocation results that do not meet the constraints are returned to the optimization scheduling model for readjustment. The readjusted results are then compared with the constraint association results again until the process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints are satisfied, thus generating the optimal heat scheduling strategy.

[0109] like Figure 4 As shown, S5: Controls the operation of heat pumps, circulating pumps, valves and boilers according to the optimal heat scheduling strategy, collects operation feedback data and process compliance results, performs rolling corrections on the optimized scheduling model based on the operation feedback data and process compliance results, and outputs correction scheduling instructions.

[0110] S5.1: Convert the optimal heat dispatch strategy into heat pump start / stop commands, circulating pump speed control commands, valve switching commands, and boiler intervention commands, and send them to the heat pump, circulating pump, valve, and boiler respectively for operation.

[0111] The optimal heat scheduling strategy extracts the heating sequence, heat replenishment sequence, heat storage sequence, and heat allocation results within each scheduling time window. The running segments of the corresponding air source heat pump nodes and high-temperature heat pump nodes in the heating sequence and heat replenishment sequence are converted into heat pump start-up and shutdown commands. The content corresponding to the circulation flow change in the heat allocation results is converted into circulation pump speed regulation commands. The content corresponding to the branch switching position in the heating sequence, heat replenishment sequence, and heat storage sequence is converted into valve switching commands. The content corresponding to the boiler node intervention time in the heat replenishment sequence is converted into boiler intervention commands. The heat pump start-up and shutdown commands are sent to the heat pumps, the circulation pump speed regulation commands are sent to the circulation pumps, the valve switching commands are sent to the valves, and the boiler intervention commands are sent to the boilers. The operation of the heat pumps, circulation pumps, valves, and boilers is controlled according to the heat pump start-up and shutdown commands, circulation pump speed regulation commands, valve switching commands, and boiler intervention commands.

[0112] S5.2: Collect temperature, flow rate, pressure, power and process compliance status of heat pumps, circulating pumps, valves and boilers during operation, and obtain operation feedback data and process compliance results.

[0113] The system reads the operating status of heat pumps, circulating pumps, valves, and boilers within the scheduling time window, and collects their temperature, flow rate, pressure, and power within the same time window. This data is then aggregated according to the scheduling time window sequence to form operational feedback data. The system reads the actual operating results of process hot water nodes, secondary sterilization nodes, and cooling nodes within the scheduling time window, and compares these results with the process hot water demand, secondary sterilization hot water demand, and cooling load demand item by item. If the actual operating results show that the process hot water temperature and flow rate meet the process hot water demand, the secondary sterilization hot water temperature and circulation flow rate meet the secondary sterilization hot water demand, and the supply and return water status during the cooling process meets the cooling load demand, the comparison result is determined as a process compliance result. Otherwise, the comparison result is determined as a process non-compliance result.

[0114] S5.3: Based on the operational feedback data and process compliance results, the deviations of the heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity and constraint parameters in the optimized scheduling model are corrected to obtain the corrected optimized scheduling model.

[0115] The temperature, flow rate, pressure, and power data from the operational feedback are compared item by item with the heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity, and constraint parameters in the optimized scheduling model to generate deviation results. When the deviation results maintain the same deviation direction for multiple consecutive scheduling time windows (e.g., three consecutive scheduling time windows), the deviation results are used as correction trigger conditions. When the temperature, flow rate, pressure, and power are lower than those in the optimized scheduling model, the heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity, and constraint parameters are reduced and tightened in the direction of deviation, and vice versa. Increase the heat pump heating capacity, boiler supplementary heating capacity, and thermal storage capacity, and relax the constraint parameters in the direction of deviation. When the process meets the standard but fails to meet the standard, prioritize correcting the heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity, and constraint parameters corresponding to the process failure. When the deviation only occurs within a single scheduling time window, keep the heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity, and constraint parameters unchanged. Write the corrected heat pump heating capacity, boiler supplementary heating capacity, thermal storage capacity, and constraint parameters back into the optimization scheduling model to obtain the corrected optimization scheduling model.

[0116] S5.4: The waste heat of different grades in the predicted supply and demand sequence and the heat pump heating capacity, boiler supplementary heating capacity and pre-peak heat storage results in the virtual energy network are re-input into the modified optimization scheduling model for collaborative optimization solution, and the modified scheduling command is output.

[0117] The predicted waste heat of different grades, the heat pump heating capacity, boiler supplementary heating capacity, and pre-peak heat storage results in the virtual energy network are re-input into the modified optimization scheduling model. Based on the heat scheduling objects, scheduling order, scheduling boundary, and scheduling objectives in the modified optimization scheduling model, the different grades of waste heat, heat pump heating capacity, boiler supplementary heating capacity, and pre-peak heat storage results are re-ordered and heat is allocated to form a modified scheduling result. The heating order, supplementary heating order, heat storage order, and heat allocation result in the modified scheduling result are converted into heat pump start-stop commands, circulating pump speed regulation commands, valve switching commands, and boiler intervention commands, and combined to obtain the modified scheduling command. The modified scheduling command is used to control the operation of heat pumps, circulating pumps, valves, and boilers within the scheduling time window.

[0118] Furthermore, the operational feedback data is compared with the heat load curve, cold load curve, waste heat output, and waste heat grade in the predicted supply and demand sequence. The deviation results obtained from the comparison are written back to the standardized thermal dataset as the basis for subsequent historical correlation segment screening, similarity matching, and trend extrapolation correction.

[0119] like Figure 2 As shown, this embodiment also provides an intelligent energy dispatching system for food production, including:

[0120] The heat supply and demand module is used to collect raw multi-source datasets, preprocess them, generate standardized thermal datasets, identify process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal datasets, and form a heat supply and demand object set.

[0121] The supply and demand sequence module is used to predict the heat load curve, cooling load curve, waste heat output and waste heat grade within a future time window based on the heat supply and demand object set and the standardized thermal data set, so as to obtain the predicted supply and demand sequence.

[0122] The scheduling model module is used to construct a virtual energy network based on the predicted supply and demand sequence, perform supply and demand mapping and constraint association on the virtual energy network, and establish an optimized scheduling model.

[0123] The pre-peak thermal storage module is used to pre-schedule the thermal storage capacity in the virtual energy network, generate pre-peak thermal storage results, and input the waste heat of different grades in the predicted supply and demand sequence, the heat pump heating capacity and boiler supplementary heating capacity in the virtual energy network, and the pre-peak thermal storage results into the optimization scheduling model for collaborative optimization and solution, and output the optimal heat scheduling strategy.

[0124] The correction scheduling module is used to control the operation of heat pumps, circulating pumps, valves and boilers according to the optimal heat scheduling strategy, and to collect operation feedback data and process compliance results. Based on the operation feedback data and process compliance results, the module performs rolling corrections to the optimized scheduling model and outputs correction scheduling instructions.

[0125] This embodiment also provides a computer device applicable to the intelligent energy scheduling method for food production, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the intelligent energy scheduling method for food production as proposed in the above embodiment.

[0126] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0127] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the energy intelligent scheduling method for food production as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0128] In summary, this invention generates a standardized thermal dataset by performing timestamp unification, missing item completion, abnormal fluctuation item removal, dimension unification, and correlation labeling on the original multi-source dataset, thereby improving the reliability of prediction and scheduling results. By classifying and labeling cooling water waste heat and plant hot air waste heat and forming a virtual energy network, the degree of relationship confusion during multi-device collaborative scheduling is reduced. By configuring the time-series correlation of heating sequence, heat replenishment sequence, and heat storage sequence, an optimized scheduling model is formed, improving the orderliness of supply and demand matching and the constraint integrity of scheduling solutions. By performing collaborative optimization again, corrected scheduling instructions are output, reducing the possibility of scheduling results becoming inaccurate during long-term operation.

[0129] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An intelligent energy scheduling method for food production, characterized in that: include, Collect raw multi-source datasets and preprocess them to generate standardized thermal datasets. Identify process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal datasets to form a set of heat supply and demand objects. The original multi-source dataset includes production planning data, energy metering data, supply and return water temperature data, flow data, pressure data, power data, and environmental monitoring data; Based on the heat supply and demand object set and standardized thermal data set, the heat load curve, cold load curve, waste heat output and waste heat grade are predicted within the future time window to obtain the predicted supply and demand sequence. A virtual energy network is constructed based on predicted supply and demand sequences. Supply and demand mapping and constraint correlation are performed on the virtual energy network, and an optimal scheduling model is established. The specific steps for constructing a virtual energy network based on predicted supply and demand sequences are as follows. Based on the process hot water demand, secondary sterilization hot water demand, and cooling load demand, process hot water node, secondary sterilization node, and cooling node are established respectively. Waste heat from cooling water and waste heat from hot air in the plant are classified and labeled by waste heat output and waste heat grade to generate waste heat nodes. Based on the relationships between preheating, temperature raising, deep heating, heat replenishment and heat storage in the food production process, air source heat pump nodes, high temperature heat pump nodes, boiler nodes and heat storage nodes are established respectively. By linking process hot water nodes, secondary sterilization nodes, cooling nodes, waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and thermal storage nodes, a virtual energy network is formed. Pre-peak thermal storage pre-scheduling is performed on the thermal storage capacity in the virtual energy network to generate pre-peak thermal storage results. Different grades of waste heat in the predicted supply and demand sequence, as well as the heat pump heating capacity, boiler supplementary heating capacity and pre-peak thermal storage results in the virtual energy network, are input into the optimization scheduling model for collaborative optimization and solution, and the optimal heat scheduling strategy is output. The operation of heat pumps, circulating pumps, valves and boilers is controlled according to the optimal heat dispatch strategy. Operational feedback data and process compliance results are collected. The optimized dispatch model is rolled over based on the operational feedback data and process compliance results, and the corrected dispatch instructions are output.

2. The intelligent energy scheduling method for food production as described in claim 1, characterized in that: The specific steps for generating the standardized thermal dataset are as follows. Unify the timestamps of the original multi-source datasets to form a unified timeline data; The data on the unified time axis are filled with missing items, abnormal fluctuation items are removed, and the units are unified to obtain standardized monitoring data. Standardized monitoring data are associated and labeled according to the production line number, process section number, equipment number, and scheduling time window in the production process to generate a standardized thermal dataset.

3. The intelligent energy scheduling method for food production as described in claim 2, characterized in that: The specific steps for forming the heat supply and demand object set are as follows: Extract the process hot water demand, secondary sterilization hot water demand, and cooling load demand in the food production process from the standardized thermal dataset to form a demand object set; Waste heat resources are obtained by extracting waste heat from cooling water and waste heat from hot air in the plant from standardized thermal datasets. The demand object set and waste heat resources are identified and aggregated to generate a heat supply and demand object set.

4. The intelligent energy scheduling method for food production as described in claim 3, characterized in that: The specific steps to obtain the predicted supply and demand sequence are as follows: The heat supply and demand object set is segmented according to the scheduling time window to generate time window supply and demand segments; By using time windows to filter historical correlation segments under the same batch, process section, and environmental conditions from the standardized thermal data set; The supply and demand segments within a time window are matched with historical related segments to generate target related segments; Based on the heat and cold load change trends of the target associated segments, the process hot water demand, secondary sterilization hot water demand, and cooling load demand are extrapolated to generate heat load curves and cold load curves. Based on the cooling water temperature difference, ambient air temperature difference, and heat pump operating status in the standardized thermal data set, waste heat assessment is performed on the heat load curve and cooling load curve to determine the waste heat output and waste heat grade. By combining the heat load curve, cooling load curve, waste heat output, and waste heat grade, a predicted supply and demand sequence is generated.

5. The intelligent energy scheduling method for food production as described in claim 4, characterized in that: The specific steps for establishing the optimized scheduling model are as follows: Read the node types and node attributes in the virtual energy network, and map process hot water nodes, secondary sterilization nodes, and cooling nodes to demand-side nodes based on node types and node attributes, and map waste heat nodes, air source heat pump nodes, high temperature heat pump nodes, boiler nodes, and thermal storage nodes to supply-side nodes. Pair demand-side nodes and supply-side nodes according to scheduling time windows to generate supply and demand mapping relationships. The heating sequence, heat replenishment sequence, and heat storage sequence in the supply and demand mapping relationship are configured with time sequence association to generate the supply and demand relationship; Based on the supply and demand relationship, configure process temperature constraints, process timing constraints, equipment capacity constraints, supply and return water pressure constraints, and boiler standby constraints, and generate constraint relationship results; By integrating the supply and demand mapping relationship, the supply and demand correlation relationship, and the constraint correlation results, the heat scheduling object, scheduling order, scheduling boundary and scheduling objective are determined, and an optimized scheduling model is formed.

6. The intelligent energy scheduling method for food production as described in claim 5, characterized in that: The specific steps for generating the pre-peak thermal storage results are as follows. Identify peak time windows based on the heat load curve in the predicted supply and demand sequence; Extract waste heat output from the scheduling time window before the peak time window, and conduct joint analysis of heat pump heating capacity, boiler supplementary heating capacity, heat storage capacity and waste heat output in the virtual energy network to identify the pre-storage time window. Peak demand during peak time windows and available heat during pre-storage time windows are shaving and valley filling to generate heat storage. The heat storage capacity is combined with the heat storage capacity in the pre-storage time window to form the pre-peak heat storage result.

7. The intelligent energy scheduling method for food production as described in claim 6, characterized in that: The specific steps for outputting the optimal heat scheduling strategy are as follows. The waste heat of different grades in the predicted supply and demand sequence is classified to generate waste heat classification results. The waste heat classification results are sequentially matched and heat is allocated through the heat pump heating capacity, boiler supplementary heating capacity and pre-peak heat storage results in the virtual energy network to generate initial scheduling results. The initial scheduling results are verified based on the constraint association results, and the optimal heat scheduling strategy is generated.

8. The intelligent energy scheduling method for food production as described in claim 7, characterized in that: The specific steps for collecting operational feedback data and process compliance results are as follows. The optimal heat dispatch strategy is converted into heat pump start / stop commands, circulating pump speed adjustment commands, valve switching commands, and boiler intervention commands, which are then sent to the heat pump, circulating pump, valve, and boiler for operation. The system collects data on temperature, flow rate, pressure, power, and process compliance during the operation of heat pumps, circulating pumps, valves, and boilers, and obtains operational feedback data and process compliance results.

9. An intelligent energy scheduling system for food production, based on the intelligent energy scheduling method for food production according to any one of claims 1 to 8, characterized in that: include, The heat supply and demand module is used to collect raw multi-source datasets, preprocess them, generate standardized thermal datasets, identify process hot water demand, secondary sterilization hot water demand, cooling load demand and waste heat resources from the standardized thermal datasets, and form a heat supply and demand object set. The supply and demand sequence module is used to predict the heat load curve, cooling load curve, waste heat output and waste heat grade within a future time window based on the heat supply and demand object set and the standardized thermal data set, so as to obtain the predicted supply and demand sequence. The scheduling model module is used to construct a virtual energy network based on the predicted supply and demand sequence, perform supply and demand mapping and constraint association on the virtual energy network, and establish an optimized scheduling model. The pre-peak thermal storage module is used to pre-schedule the thermal storage capacity in the virtual energy network, generate pre-peak thermal storage results, and input the waste heat of different grades in the predicted supply and demand sequence, the heat pump heating capacity and boiler supplementary heating capacity in the virtual energy network, and the pre-peak thermal storage results into the optimization scheduling model for collaborative optimization and solution, and output the optimal heat scheduling strategy. The correction scheduling module is used to control the operation of heat pumps, circulating pumps, valves and boilers according to the optimal heat scheduling strategy, and to collect operation feedback data and process compliance results. Based on the operation feedback data and process compliance results, the module performs rolling corrections to the optimized scheduling model and outputs correction scheduling instructions.

Images