Coconut product production line control method and system

Abstract

The invention relates to the technical field of food processing automation, in particular to a coconut product production line control method and system. The method comprises the steps of firstly obtaining production plan data issued by a manufacturing execution system; inputting the data into a pre-trained energy consumption prediction model, and generating a predicted energy consumption sequence of each production line in a planned time period; polymerizing all the sequences to form a total energy demand load curve of the production system in a future period; then, based on the load curve and the real-time energy supply state, an active scheduling instruction containing a high-energy-consumption process time sequence optimization scheme is generated; and finally, generating and issuing a time sequence control instruction to each production line according to the instruction, and dynamically adjusting the actual starting time of the high-energy-consumption process, so that the real-time energy consumption load of the system is matched with the optimization scheme. According to the method, deep collaboration of the production plan and energy management is realized, and the problems of energy consumption prediction loss and scheduling passivity caused by disjunction of the production plan and the energy management are solved, so that the energy efficiency and the execution flexibility of the production plan are improved.

Description

Technical Field

[0001] This invention relates to the field of food processing automation technology, and in particular to a production line control method and system for coconut products. Background Technology

[0002] In the field of large-scale coconut product processing, to ensure the process independence of different products such as frozen coconut cream, raw coconut milk, and coconut water, multiple dedicated production lines are usually set up in parallel. Each production line is equipped with an independent production control system, which is responsible for managing its internal process logic and process parameters.

[0003] Currently, factories typically deploy Manufacturing Execution Systems (MES) at the production management level to develop detailed production plans; simultaneously, they deploy Energy Management Systems (EMS) at the energy management level to monitor real-time energy consumption. However, these two systems are usually independent in terms of function and data. The production control system only executes instructions from the MES, while the energy management system passively collects historical and real-time energy consumption data. This architecture leads to a critical technological bottleneck: the energy management center cannot obtain, in advance and accurately, the foreseeable energy demand load curves based on the specific production plans of the three production lines for a future period.

[0004] Due to the lack of this predictive information, energy dispatch is in a passive and lagging state. On the one hand, it is impossible to proactively coordinate the start-up and shutdown sequence of high-energy-consuming equipment, such as sterilization tanks and concentration units, to achieve peak shaving and valley filling, which may result in high energy costs. On the other hand, when faced with total energy quotas, it is difficult to dynamically and precisely adjust the production sequence and progress of each production line based on the principle of optimal overall benefits to ensure the delivery of key orders. Summary of the Invention

[0005] This invention provides a production line control method and system for coconut products, which solves the technical problem in the prior art where the disconnect between production control and energy management leads to the inability to achieve planned energy demand forecasting and proactive scheduling, resulting in low energy efficiency and inflexible production plan execution.

[0006] The first aspect of this invention provides a production line control method for coconut products, applied to a production system comprising at least two independent production lines, the method comprising:

[0007] Obtain production plan data issued by the Manufacturing Execution System;

[0008] The production plan data is input into a pre-trained energy consumption prediction model, which generates a predicted energy consumption sequence for each production line during the corresponding planned production period.

[0009] Aggregate all the predicted energy consumption sequences to generate the total energy demand load curve of the production system within a preset future period;

[0010] Based on the total energy demand load curve and real-time energy supply status, active scheduling instructions are fed back; the active scheduling instructions include an optimization scheme for the start-up sequence of high energy-consuming processes in each of the production lines.

[0011] Based on the active scheduling instruction, corresponding timing control instructions are generated and sent to the control systems of each production line to dynamically adjust the actual start-up time of high-energy-consuming processes in each production line, so that the real-time energy consumption load of the production system matches the optimization scheme.

[0012] Optionally, the energy consumption prediction model is pre-trained based on the historical production data of the production system; the historical production data includes at least: time-series power data, raw material input and output data for each process under different product categories.

[0013] Optionally, the predicted energy consumption sequence includes an energy consumption subsequence for a target high-energy-consuming process; the target high-energy-consuming process includes at least one of an instantaneous sterilization process, a high-temperature sterilization process, a homogenization process, and an evaporation and concentration process.

[0014] Optionally, the step of aggregating all the predicted energy consumption sequences to generate the total energy demand load curve of the production system within a preset future period includes:

[0015] Based on a preset future period, an initial reference time axis with a uniform time interval is set;

[0016] The predicted energy consumption sequence corresponding to each production line is mapped onto the initial reference time axis according to the corresponding planned production period to generate a target reference time axis; wherein, for the idle period outside the planned production period, the energy consumption value on the target reference time axis is filled with zero;

[0017] The total system load prediction value corresponding to each time point on the target reference time axis is calculated by summing up all the predicted energy consumption values ​​at each time point.

[0018] Connect the predicted total system load values ​​at all time points on the target reference time axis to generate the total energy demand load curve corresponding to the production system.

[0019] Optionally, the step of feeding back active dispatch instructions based on the total energy demand load curve and real-time energy supply status includes:

[0020] Obtain externally inputted energy constraints and production optimization objectives; wherein, the energy constraints include the upper limit of contracted electricity capacity or total energy consumption limit within a future set period; the production optimization objectives include ensuring timely delivery of key orders;

[0021] The total energy demand load curve is compared with the energy constraints to determine the load conflict period;

[0022] Based on the traceable load composition information in the total energy demand load curve, identify one or more target high-energy-consuming processes that contribute the most to the load during the load conflict period.

[0023] For each of the target high-energy-consuming processes, within the time flexibility window allowed by the process of the target high-energy-consuming process, calculate the adjustable amount of time that the target high-energy-consuming process can be delayed or started earlier.

[0024] When multiple target high-energy-consuming processes are of the same type and their planned time periods overlap within the load conflict period, an execution sequence table corresponding to the production system is constructed on the premise of ensuring the continuity of core production batches of each production line.

[0025] When a conflict cannot be completely resolved by timing adjustments alone, a priority list corresponding to the production system is constructed based on the preset product category priority; wherein, the preset product category priority is set as follows: coconut water production line is higher than raw coconut milk production line, and raw coconut milk production line is higher than frozen coconut milk production line.

[0026] The adjustable time amount, the execution sequence list, and the priority list are combined and conflict checked to generate candidate scheduling instruction schemes.

[0027] The effectiveness of the candidate scheduling instructions in mitigating load conflicts and the disturbances to the production plan are evaluated. Based on the production optimization objectives, the optimal solution is selected and formatted as an active scheduling instruction output.

[0028] Optionally, the step of generating and issuing corresponding timing control instructions to the control systems of each production line according to the active scheduling instruction includes:

[0029] The active scheduling instruction is parsed to extract the adjustable time amount, execution sequence list and priority list associated with each production line. Based on the execution sequence list and priority list, the adjustable time amount is allocated according to the production line identifier to generate independent control tasks corresponding to each production line.

[0030] Each independent control task is verified to meet the preset constraints of the corresponding production line, and verification data is generated. The preset constraints include: the material waiting time caused by process adjustment does not exceed its maximum allowable shelf life, and the production line that is subsequently downgraded in the priority list is in a state where it is safe to reduce load.

[0031] When the verification data passes the verification, the planned start timestamp of the corresponding high-energy-consuming process on the production line is calculated based on the adjustment time of the independent control task corresponding to the verification data and its sorting position in the execution sequence list, and a timing control instruction containing the planned start timestamp is generated.

[0032] The timing control commands corresponding to each production line are sent to the corresponding production line control system, and the actual start-up time of high-energy-consuming processes is monitored.

[0033] When the deviation between the actual start time and the corresponding planned start timestamp exceeds a preset threshold, automatic fine-tuning of the cycle time of subsequent processes is triggered.

[0034] A second aspect of the present invention provides a production line control system for coconut products, applied to a production system comprising at least two independent production lines, the system comprising:

[0035] The production planning data acquisition module is used to acquire production planning data issued by the manufacturing execution system.

[0036] The energy consumption prediction module is used to input the production plan data into a pre-trained energy consumption prediction model, and the energy consumption prediction model generates a predicted energy consumption sequence for each production line during the corresponding planned production period.

[0037] The load curve aggregation module is used to aggregate all the predicted energy consumption sequences to generate the total energy demand load curve of the production system in a preset future period.

[0038] An active scheduling instruction generation module is used to feed back active scheduling instructions based on the total energy demand load curve and real-time energy supply status; the active scheduling instructions include an optimization scheme for the start-up sequence of high energy-consuming processes in each of the production lines.

[0039] The timing control instruction execution module is used to generate and issue corresponding timing control instructions to the control systems of each production line according to the active scheduling instructions, so as to dynamically adjust the actual start time of high energy-consuming processes in each production line, so as to match the real-time energy consumption load of the production system with the optimization scheme.

[0040] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the production line control method for coconut products as described above.

[0041] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the production line control method for coconut products as described above.

[0042] The fifth aspect of the present invention provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein, when the program instructions are executed by a computer, the computer performs the production line control method for coconut products as described above.

[0043] As can be seen from the above technical solutions, the present invention has the following advantages:

[0044] This invention acquires production plan data from a manufacturing execution system (MES), inputs it into a pre-trained energy consumption prediction model to predict energy consumption, generates predicted energy consumption sequences for each production line's planned production period, and then aggregates all predicted energy consumption sequences to obtain a total energy demand load curve for the production system's preset future cycle. Based on this curve and real-time energy supply status feedback, it includes proactive scheduling instructions with optimized start-up timing schemes for high-energy-consuming processes. Finally, based on these proactive scheduling instructions, it generates and issues timing control instructions to the control systems of each production line, dynamically adjusting the actual start-up times of high-energy-consuming processes to achieve precise matching between the production system's real-time energy load and the optimized scheme, effectively connecting the entire process of production control and energy management. This solution solves the technical problem in existing technologies where the disconnect between production control and energy management leads to the inability to achieve planned energy demand prediction and proactive scheduling, resulting in low energy efficiency and inflexible production plan execution.

[0045] This invention deeply integrates production planning with energy consumption forecasting, enabling energy demand prediction to be based on actual production tasks. This breaks away from the isolated model of traditional energy management detached from production planning, allowing for advance prediction of energy consumption distribution across production lines during planned production periods. This avoids energy waste or supply shortages caused by inaccurate energy demand forecasts, significantly improving energy efficiency. Specifically, the aggregated total energy demand load curve clearly shows the peak and trough energy consumption within a preset future period, providing accurate data support for proactive scheduling and enabling energy allocation to adapt to production needs in advance. Through a proactive scheduling mechanism based on the total energy demand load curve and real-time energy supply status, dynamic adaptation of energy management to the production process is achieved. This changes the passive mode of traditional production control that only focuses on completing production tasks while ignoring energy constraints. By optimizing the start-up sequence of high-energy-consuming processes, energy supply and demand conflicts can be avoided while ensuring the progress of production plans, improving the flexibility of production plan execution. For example, when the total energy demand load curve shows a risk of peak load exceeding real-time energy supply capacity during a certain period, proactive scheduling commands can adjust the start-up sequence of high-energy-consuming processes, smoothing the peak load to a reasonable range and preventing production interruptions due to insufficient energy. From the perspective of the collaborative closed loop of production control and energy management, the precise adjustment of the actual start time of high-energy-consuming processes by the timing control instructions ensures that the optimization scheme of proactive scheduling can be effectively implemented, and achieves precise matching between the real-time energy consumption load of the production system and the optimization scheme. This not only solves the problem of low energy efficiency of existing technologies, but also replaces passive response with proactive scheduling, enabling production plans to be flexibly adjusted according to energy status, ensuring a dual improvement in production efficiency and energy utilization efficiency. At the same time, it is suitable for the production scenario of multiple production lines of coconut products operating in parallel, providing strong support for the stable production of multiple categories of coconut products. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0047] Figure 1 A flowchart illustrating the steps of a production line control method for coconut products provided in an embodiment of the present invention;

[0048] Figure 2 A structural block diagram of a production line control system for coconut products provided in an embodiment of the present invention;

[0049] Figure 3 This is a structural block diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0050] This invention provides a production line control method and system for coconut products, which solves the technical problem in the prior art where the disconnect between production control and energy management leads to the inability to achieve planned energy demand forecasting and proactive scheduling, resulting in low energy efficiency and inflexible production plan execution.

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that in the optional embodiments of the present invention, the object information and other related data involved require the permission or consent of the object when the embodiments of the present invention are applied to specific products or technologies, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. That is to say, if the embodiments of the present invention involve data related to the object, it needs to be obtained with the authorization and consent of the object, the authorization and consent of the relevant departments, and in compliance with the relevant laws, regulations, and standards of the country and region. If personal information is involved in the embodiments, the acquisition of all personal information requires the consent of the individual. If sensitive information is involved, the separate consent of the information subject is required, and the embodiments also need to be implemented with the authorization and consent of the object.

[0052] Please see Figure 1 , Figure 1 A flowchart illustrating the steps of a production line control method for coconut products provided in an embodiment of the present invention.

[0053] This invention provides a production line control method for coconut products, applicable to a production system comprising at least two independent production lines. Specifically, this system is an automated production cluster for processing different types of coconut products. A typical configuration includes three core independent production lines: a coconut water production line, a raw coconut milk production line, and a frozen coconut cream production line. Each production line has a complete process chain from raw material pretreatment to finished product packaging, and operates independently (i.e., material transfer, equipment control, and process cycle time of each production line do not directly interfere with each other). Simultaneously, the production system integrates a unified energy data acquisition terminal, an edge computing server, and an industrial communication network. The energy data acquisition terminal is deployed next to high-energy-consuming equipment (such as instant sterilizers and evaporation concentration devices) on each production line to collect real-time data on equipment power and energy consumption. The edge computing server serves as the core computing platform for the entire control method, incorporating the energy consumption prediction model and scheduling decision algorithm of this application, and possessing data interaction capabilities with the PLC control system and Manufacturing Execution System (MES) of each production line. The industrial communication network uses a combination of PROFIBUS-DP bus and TCP / IP protocol to ensure real-time transmission of production plan data, control commands, and energy consumption data, with transmission delay controlled within 1 second, meeting the real-time control requirements of the production line. It should be noted that this application does not impose a unique limitation on the number of production lines. Any scenario involving at least two independent production lines and overlapping energy consumption conflicts in high-energy-consuming processes falls within the scope of this application. For example, a dual-production line system consisting only of a coconut water production line and a raw coconut milk production line is also applicable to this method.

[0054] The method includes:

[0055] Step 101: Obtain the production plan data issued by the Manufacturing Execution System.

[0056] In this embodiment of the invention, the production plan data is a structured production task list generated by the manufacturing execution system according to the order requirements. It is not generalized production information. It contains at least the following core dimensions, and the data of each dimension are precisely matched with the production characteristics and energy consumption prediction requirements of coconut products: (1) Product category information: the type of coconut products to be produced on each production line is clearly defined. There are significant differences in the process links and high energy consumption process parameters corresponding to different categories. For example, coconut water needs to be sterilized at low temperature and frozen coconut milk needs to go through an additional freezing process; (2) Batch size and production time information: it includes the raw material input and finished product output of each batch on each production line, as well as the planned production time corresponding to each batch. This information is used to determine the start time range of high energy consumption processes and the time dimension of energy consumption prediction; (3) Order priority information: the urgency of the orders corresponding to each production batch is marked. Among them, the urgent orders for coconut water need to be delivered first. This information complements the product category priority in the subsequent scheduling optimization and jointly supports the realization of the production optimization goal; (4) Process parameter requirements: the core process parameters of each high energy consumption process are clearly defined. These parameters directly affect the energy consumption of the equipment and are important input features of the energy consumption prediction model.

[0057] Data acquisition is achieved through a production plan data acquisition module built into the edge computing server. This module establishes a stable communication connection with the Manufacturing Execution System (MES) via an industrial Ethernet interface and uses the OPC UA communication protocol for data transmission. The OPC UA communication protocol features cross-platform compatibility and high reliability, ensuring the integrity of production plan data transmission in complex industrial environments. The specific acquisition process is as follows: the edge computing server sends data request commands to the Manufacturing Execution System at a preset cycle (preferably every 5 minutes). After receiving the request, the Manufacturing Execution System packages and returns the currently confirmed production plan data in a preset JSON format. If there are changes to the production plan (such as adding urgent orders or adjusting production periods), the Manufacturing Execution System will proactively push the changed data to the edge computing server to ensure that the production plan data acquired in step 101 is real-time and accurate.

[0058] Step 102: Input the production plan data into the pre-trained energy consumption prediction model, and the energy consumption prediction model will generate the predicted energy consumption sequence for each production line during the corresponding planned production period.

[0059] In this embodiment of the invention, the production plan data is first preprocessed to ensure that the data format and dimensions input to the model are consistent with the training standards. Eight key features, including product category, raw material input, planned process duration, and core process parameters, are extracted from the processed production plan data. Then, a Min-Max normalization method is used to map features of different magnitudes, such as raw material input and process parameters, to the [0, 1] interval. The normalization formula remains consistent with that used during model training to avoid the impact of data magnitude differences on computational accuracy. The preprocessed data is synchronously associated with the corresponding production line identifier and planned production period to ensure compatibility with the subsequent mapping reference time axis. Next, the model prediction computation is initiated, relying on a pre-trained LSTM model deployed on an edge computing server to perform core computations. The model calls upon the historical energy consumption data of coconut product production from the past six months stored locally, and combines it with the preprocessed feature vectors to accurately capture the temporal patterns of energy consumption changes in each process with production parameters, focusing on fitting the energy consumption fluctuation characteristics of high-energy-consuming processes such as instantaneous sterilization and evaporation concentration. These processes have a high energy consumption ratio and are greatly affected by the characteristics of coconut raw materials. The model has learned from historical data to adapt to the energy consumption differences caused by the moisture content of coconuts and the raw material loss rate in different seasons. The calculation process uses the corresponding planned production period as the time axis and adopts a sampling interval of 1 minute per time point. Finally, a predicted energy consumption sequence is generated and the process dimension is decomposed. The model outputs the predicted energy consumption values ​​for each time point in chronological order, forming an initial predicted energy consumption sequence that is perfectly aligned with the planned production period of the corresponding production line. The time span of the sequence is precisely synchronized with the planned production period.

[0060] Furthermore, the energy consumption prediction model is pre-trained based on historical production data of the production system; the historical production data includes at least: time-series power data of each process under different product categories, raw material input and output data.

[0061] In this embodiment of the invention, historical production data are all derived from the coconut product production system applicable to this application, ensuring data consistency with actual application scenarios. Specifically, the time-series power data for each process under different product categories refers to the real-time power data of the corresponding equipment in each process of different product categories such as coconut water, raw coconut milk, and frozen coconut cream, collected at a sampling frequency of 1 minute per data point, forming a time-based sequence of data (e.g., in the production of freshly squeezed coconut water, the power sequence of the instantaneous sterilization process from 8:00 to 8:10 is [350kW, 352kW, ..., 349kW]). Raw material input and output data refer to the actual input weight of raw materials such as coconut meat and coconut juice and the actual output quantity of finished products for each production batch, and batch association identifiers are established with the corresponding process time-series power data. The aforementioned historical production data is synchronously collected through the energy data acquisition terminal integrated into the production system (deployed next to the equipment in each process) and the production line PLC control system, and after format standardization processing (unifying data units and supplementing missing data), it is stored in the local database of the edge computing server.

[0062] In this embodiment, the energy consumption prediction model adopts a Long Short-Term Memory (LSTM) neural network model. This model has the advantage of capturing long-term dependencies in time-series data and is adapted to the characteristics of energy consumption changes over time. The specific training process is as follows: First, data preprocessing: Valid data from the past 6 months is selected from historical production data and divided into training, validation, and test sets in a 7:2:1 ratio. Data normalization is also performed to avoid the impact of differences in data volume on training results. Second, model structure construction: The input layer dimension matches the feature dimensions of the production plan data (including 8 features such as product category code, raw material input, planned process duration, and process parameters). Three hidden layers are set with 64, 128, and 64 neurons respectively. The output layer is the energy consumption sequence of the corresponding production line during the planned production period. Third, training and optimization: The root mean square error (RMSE) < 5% is used as the training convergence criterion. The Adam optimization algorithm is used to adjust the model parameters, and the model's generalization ability is monitored in real time through the validation set to avoid overfitting. The fourth step is model validation and deployment. The model prediction accuracy is validated using a test set. When the energy consumption prediction error of different product categories meets the requirement of RMSE < 5%, the model training is completed. The trained model is then deployed to the energy consumption prediction module of the edge computing server to receive the production plan data output in step 101 and perform predictions.

[0063] Furthermore, the predicted energy consumption sequence includes an energy consumption subsequence targeting a high-energy-consuming process; the high-energy-consuming process includes at least one of an instantaneous sterilization process, a high-temperature sterilization process, a homogenization process, and an evaporation and concentration process.

[0064] In this embodiment of the invention, the selection of target high-energy-consuming processes is not arbitrary, but determined based on the energy consumption ratio analysis results of the coconut product production system of this application. Statistical analysis of the production system's historical production data for the past six months shows that the energy consumption of the instantaneous sterilization process, high-temperature sterilization process, homogenization process, and evaporation and concentration process accounts for 22%, 18%, 15%, and 45% of the total energy consumption of each production line, respectively, totaling 100%. These are the core processes affecting the energy consumption of the production lines and the total load of the system. Furthermore, these processes exhibit significant energy consumption concentration characteristics (e.g., the instantaneous power of the evaporation and concentration process can reach over 500kW after startup, far exceeding other processes), and are also the main cause of energy consumption conflicts across multiple production lines. Therefore, they are defined as target high-energy-consuming processes, allowing for focused optimization of subsequent scheduling and improving scheduling efficiency. It should be noted that this application does not impose a unique limit on the number of target high-energy-consuming processes; any process including at least one of the aforementioned core high-energy-consuming processes falls within the scope of protection of this application, adapting to the differences in process configurations of different production lines.

[0065] The predicted energy consumption sequence is a complete energy consumption data sequence with time as its axis, and its time span is completely consistent with the planned production period obtained in step 101. This sequence is not a single, unified data set, but rather broken down into multiple energy consumption sub-sequences according to the process dimension. The core component is the energy consumption sub-sequence of the target high-energy-consuming process, while also including energy consumption sub-sequences of low-energy-consuming processes such as raw material pretreatment and finished product packaging (the energy consumption sub-sequences of low-energy-consuming processes are used to ensure the integrity of the sequence, and their energy consumption proportion is low, so they are not included in the core objects of subsequent scheduling optimization). For example, the predicted energy consumption sequence of the coconut water production line from 8:00 to 10:30 includes "energy consumption sub-sequence of instant sterilization process (8:10-8:20), energy consumption sub-sequence of evaporation and concentration process (8:30-10:00), and energy consumption sub-sequence of low-energy-consuming processes (8:00-8:10, 8:20-8:30, 10:00-10:30)," and the time interval of each sub-sequence is precisely matched with the planned execution period of the corresponding process.

[0066] It is important to emphasize that the energy consumption prediction model and the generation of the predicted energy consumption sequence in this step are not simply a general application of industrial energy consumption prediction technology, but rather a customized optimization for the specific characteristics of the coconut product production system. The model training data focuses on the process characteristics of different categories of coconut products, and the prediction results are broken down to adapt to subsequent scheduling needs.

[0067] Step 103: Aggregate all predicted energy consumption sequences to generate the total energy demand load curve of the production system within a preset future period.

[0068] Furthermore, step 103 may include the following sub-steps:

[0069] S11. Based on a preset future period, set an initial reference time axis with a uniform time interval.

[0070] It should be noted that the preset future period is not arbitrarily set, but is precisely matched with the batch production time period in the production plan data obtained in step 101, and determined in combination with the scheduling cycle requirements of the production system. The preferred values ​​are 24 hours or 48 hours. For example, if the production plan is a daily batch plan (e.g., each production line completes 3-4 production batches per day), the preset future period is set to 24 hours; if there are long-cycle production batches spanning multiple days (e.g., a batch of frozen coconut milk has a production time of 15 hours, spanning from 16:00 on the current day to 7:00 the next day), the preset future period is set to 48 hours to ensure complete coverage of all planned production time periods. The time interval is consistent with the sampling interval of the predicted energy consumption sequence in step 102, set to 1 minute per time point. This ensures that the time accuracy matches the energy consumption fluctuation characteristics of high-energy-consuming processes (e.g., the power can rise to peak within 1 minute when the instantaneous sterilization process starts), while avoiding data redundancy and increased computational load on the edge computing server due to excessively short time intervals.

[0071] In this embodiment of the invention, the starting time of a preset future cycle is taken as the origin (e.g., the starting time of a 24-hour cycle is 0:00 of the day), and time nodes are divided at 1-minute intervals to form a continuous time axis containing 1440 time points (24 hours × 60 minutes) or 2880 time points (48 hours × 60 minutes), thus obtaining the initial reference time axis.

[0072] S12. Map the predicted energy consumption sequence corresponding to each production line to the initial reference time axis according to the corresponding planned production period to generate the target reference time axis; wherein, for the idle period outside the planned production period, fill its energy consumption value on the target reference time axis with zero.

[0073] In this embodiment of the invention, the planned production periods corresponding to each production line are extracted from the production plan data in step 101. Then, the predicted energy consumption sequences of each production line generated in step 102 (with time axes consistent with their own planned production periods) are mapped one-to-one to the initial reference time axis set in S11. For example, the energy consumption value corresponding to 8:00 in the predicted energy consumption sequence of the coconut water production line is directly mapped to the 8:00 time point on the initial reference time axis; the energy consumption value corresponding to 9:00 in the predicted energy consumption sequence of the raw coconut milk production line is mapped to the 9:00 time point on the initial reference time axis. For idle periods outside the planned production periods (such as the coconut water production line having no production plan from 0:00 to 8:00 and from 10:30 to 24:00), the energy consumption values ​​at the corresponding time points on the initial reference time axis are filled with zero, ultimately generating a target reference time axis that covers all time points of the preset future cycle and includes energy consumption data of all production lines.

[0074] S13. Accumulate all the predicted energy consumption values ​​corresponding to each time point on the target reference time axis to calculate the predicted total system load value corresponding to the time point.

[0075] In this embodiment of the invention, for each time point on the target reference time axis (e.g., 8:05), the predicted energy consumption values ​​corresponding to all production lines at that time point are extracted (including the actual predicted energy consumption values ​​of production lines in production and zero values ​​of production lines in idle state). An accumulation calculation is performed by the load curve aggregation module of the edge computing server to obtain the predicted total system load value for that time point. For example, at 8:05, the coconut water production line is in production with a predicted energy consumption of 350kW, the raw coconut milk production line is in idle state with a predicted energy consumption of 0kW, and the frozen coconut milk production line is in production with a predicted energy consumption of 480kW. Therefore, the predicted total system load value for that time point is 350kW + 0kW + 480kW = 830kW. S14. Connect the predicted total system load values ​​of all time points on the target reference time axis to generate the total energy demand load curve corresponding to the production system.

[0076] In this embodiment of the invention, the edge computing server's plotting algorithm uses the predicted total system load at all time points on the target reference time axis as the vertical axis and the time points as the horizontal axis, connecting the points in chronological order to generate a smooth total energy demand load curve. This curve clearly shows the peak, valley, and duration of the production system's energy load within a preset future period. For example, it can intuitively identify 9:30-10:00 as the peak load period with a peak load of 1200kW; and 14:00-16:00 as the valley load period with a valley load of 300kW.

[0077] It is particularly important to emphasize that the aggregation strategy in step 103 is customized and optimized for the characteristics of the coconut product production system: First, the time parameters (preset future cycle, unified time interval) are adapted to the batch duration of coconut product production and the energy consumption fluctuation characteristics of high-energy-consuming processes; second, the aggregation results are accurately associated with the load of core high-energy-consuming processes, adapting to the needs of subsequent coconut product production line scheduling optimization; third, through the rules of time alignment and empty zero filling, the problem of energy consumption aggregation caused by the staggered production of multiple batches of coconut products on multiple production lines is solved.

[0078] Step 104: Based on the total energy demand load curve and real-time energy supply status, provide feedback on proactive dispatch instructions.

[0079] Furthermore, step 104 may include the following sub-steps:

[0080] S21. Obtain external input energy constraints and production optimization objectives; wherein, energy constraints include the upper limit of contracted electricity capacity or total energy consumption limit within a future set period; and production optimization objectives include ensuring timely delivery of key orders.

[0081] In this embodiment of the invention, the energy constraints are obtained by connecting the communication module of the edge computing server to the Enterprise Energy Management System (EMS) to obtain in real time the upper limit of contracted electricity capacity for a future set period (e.g., a peak limit of 1200kW signed between the enterprise and the power supply department) or the total energy consumption limit (e.g., 15000kWh per day). The production optimization target is obtained by extracting order priority information (e.g., coconut water orders marked "urgent") from the production plan data obtained in step 101, and determining specific standards to ensure the timely delivery of key orders (e.g., delivery delay time of key orders ≤ 0 minutes) in conjunction with the enterprise's production management rules.

[0082] S22. Compare the total energy demand load curve with the energy constraints to determine the load conflict period.

[0083] In this embodiment of the invention, the predicted total system load at each time point in the total energy demand load curve is compared one by one with the energy constraints (such as the contracted electricity capacity limit of 1200kW) obtained in S21. The comparison logic is as follows: if the predicted total system load at a certain time point is greater than the energy constraint limit, then that time point is marked as a conflict time point. Subsequently, the intervals of consecutive conflict time points are statistically analyzed. If the predicted total load within a certain continuous time interval (≥5 minutes, the duration is determined by the energy consumption continuity characteristics of the high-energy-consuming process; historical data shows that the peak load duration of the target high-energy-consuming process is ≥5 minutes; if the duration is set too short, instantaneous fluctuations will be mistakenly judged as conflicts) exceeds the constraint limit, then the interval is defined as a load conflict period.

[0084] For example, the total energy demand load curve shows that the load value between 9:30 and 10:10 is consistently between 1350kW and 1420kW (the load value at each time point during this period is calculated by aggregation in step 103), exceeding the contracted upper limit of 1200kW. By traversing all time point data within this period, it is confirmed that there are conflict time points for 40 consecutive minutes. Therefore, 9:30-10:10 is determined to be a load conflict period. This logic ensures the accuracy of conflict identification while avoiding misjudgment of conflicts due to instantaneous load fluctuations, adapting to the operating characteristics of high-energy-consuming processes in coconut product manufacturing.

[0085] S23. Based on the traceable load composition information in the total energy demand load curve, identify one or more target high-energy-consuming processes that contribute the most to the load during periods of load conflict.

[0086] In this embodiment of the invention, the load value at each time point of the total energy demand load curve is the sum of the predicted energy consumption values ​​of each production line at the corresponding time point. The predicted energy consumption value of each production line can be further decomposed into the target high-energy-consuming process energy consumption subsequence and the low-energy-consuming process energy consumption subsequence defined in step 102. Based on this tracing relationship, the total load value during the conflict period is decomposed in reverse through the load curve aggregation module of the edge computing server. The specific decomposition process is as follows: extract the target high-energy-consuming process energy consumption subsequence values ​​for each production line at each time point within the conflict period; calculate the energy consumption proportion of a single target high-energy-consuming process at that time point (energy consumption value of a single process ÷ total load value at that time point × 100%); then average the proportions at each time point within the conflict period to obtain the average load contribution proportion of the target high-energy-consuming process during the conflict period. By comparing the average contribution proportions of each target high-energy-consuming process, the process with the largest load contribution is determined.

[0087] For example, during the conflict period from 9:30 to 10:10, it was found that the energy consumption of the evaporation and concentration process in the coconut water production line was 580kW-620kW at each time point, and the total load at the corresponding time point was 1350kW-1420kW, with an average proportion of 45%. The energy consumption of the evaporation and concentration process in the raw coconut milk production line was 510kW-550kW at each time point, with an average proportion of 38%. The average proportion of other processes was less than 10%. Therefore, these two high-energy-consuming processes of the same type were identified as the core of the conflict, that is, one or more high-energy-consuming processes that contributed the most to the load during the load conflict period.

[0088] S24. For each target high-energy-consuming process, within the time flexibility window allowed by the process of the target high-energy-consuming process, calculate the adjustable amount of time that the target high-energy-consuming process can be delayed or started earlier.

[0089] In this embodiment of the invention, the time flexibility window for the target high-energy-consuming process is: ±10 minutes for instant sterilization, ±20 minutes for high-temperature sterilization, ±30 minutes for homogenization, and ±30 minutes for evaporation and concentration. This parameter was verified through multiple sets of process experiments: Production experiments were conducted on each target high-energy-consuming process for different product categories at different time offsets. Finished product quality indicators (such as microbial content, moisture content, and flavor value) were tested. The maximum time offset by which the finished product indicators met the GB / T31325-2014 standard was determined, which is the flexibility window corresponding to the target high-energy-consuming process (e.g., if the evaporation and concentration process is delayed by 30 minutes, the coconut milk moisture content is still controlled within the acceptable range of 25% ± 1%; if delayed by 35 minutes, the moisture content exceeds the range, therefore the flexibility window is set to ±30 minutes). Secondly, the calculation logic for the adjustable time is as follows: First, determine the original planned start time of the target high-energy-consuming process (…). (derived from the process scheduling data of the production plan in step 101), and the process flexibility window ([ - , + ], For flexible window duration (e.g., 30 minutes) and load conflict periods ([ , [e.g., 9:30-10:10); the second step is to calculate the overlap between the flexible window and the conflict period: if During the conflict period, the overlapping interval is [ , + ]and[ , The intersection (delay direction) of ] or [ - , ]and[ , The intersection of ] (advance direction); the third step is to adjust the time amount = the duration of the overlapping interval, and to ensure the end time of the adjusted process ( +Δt+process time, where Δt is an adjustable time amount, and the process time comes from the process standard document) does not affect the normal start of subsequent processes (the start time of subsequent processes comes from the production planning and scheduling data).

[0090] For example: the original planned start time of the evaporation and concentration process in the coconut water production line. =9:40 (flexible window ±30 minutes, i.e. [9:10, 10:10]), conflicting time period [9:30, 10:10], overlapping interval [9:40, 10:10] (delay direction), duration 30 minutes; however, the subsequent freezing process is planned to start at 10:30, and the evaporation and concentration process takes 30 minutes. If the start is delayed by 30 minutes, the end time will be 10:40, which exceeds the start time of the subsequent process. Therefore, the adjustable time is corrected to 20 minutes (delayed to start at 10:00, end time 10:30, connecting with the subsequent process), that is, the final adjustable time is a delay of 20 minutes; the overlapping interval of the advance direction is [9:30, 9:40], duration 10 minutes. It has been confirmed that starting 10 minutes earlier (9:30) does not conflict with the preceding homogenization process (8:50-9:25), so starting 10 minutes earlier can also be chosen.

[0091] S25. When multiple high-energy-consuming processes are of the same type and their planned time periods overlap during load conflict periods, an execution sequence list corresponding to the production system is constructed on the premise of ensuring the continuity of core production batches of each production line.

[0092] In this embodiment of the invention, high-energy-consuming processes of the same type refer to core energy-consuming processes of the same type (such as the evaporation and concentration processes of two production lines). Core production batch continuity means that adjustments should not cause batch interruptions in the production line (e.g., after adjusting the evaporation and concentration process of a frozen coconut milk batch, the subsequent freezing process must be seamlessly connected, and the material waiting time should not exceed the maximum allowable shelf life of 4 hours). The execution sequence list is constructed as follows: based on the adjustable time of each similar process, it is sorted according to the principle of staggered start-up, and the start-up time after adjustment is clearly defined. For example, the evaporation and concentration process of the coconut water production line is adjusted to start at 10:00, and the raw coconut milk production line is adjusted to start at 10:10, forming a staggered sequence of "10:00-10:30 (coconut water), 10:10-10:40 (raw coconut milk)" to avoid load superposition.

[0093] S26. When a conflict cannot be completely resolved by timing adjustment alone, a priority list corresponding to the production system is constructed based on the preset product category priority. The preset product category priority is set as follows: coconut water production line is higher than raw coconut milk production line, and raw coconut milk production line is higher than frozen coconut milk production line.

[0094] In this embodiment of the invention, the priority list is constructed based on the comprehensive scoring results of three sets of quantitative indicators. The core construction logic is as follows: First, the three sets of core quantitative indicators affecting priority are determined, namely, product shelf life, process error tolerance, and energy consumption sensitivity. Second, the weight ratio of each indicator is calculated using the analytic hierarchy process (AHP) to clarify the degree of influence of each indicator on priority. Specifically, the weights are set as follows: shelf life 0.5, process error tolerance 0.3, and energy consumption sensitivity 0.2. Subsequently, standardized scoring rules for each indicator are formulated, using a 10-point scoring system. The higher the score, the higher the priority of that indicator. Specifically, the shelf life indicator score is negatively correlated with the shelf life duration (the shorter the shelf life, the higher the score); the process error tolerance indicator score is negatively correlated with the finished product defect rate after process adjustment (the lower the defect rate, the higher the score); and the energy consumption sensitivity indicator score is negatively correlated with the degree of impact of the adjustment on the total load (the smaller the impact, the higher the score). Then, using the formula "Comprehensive Score = Shelf Life Weight × Shelf Life Score + Process Fault Tolerance Weight × Fault Tolerance Score + Energy Consumption Sensitivity Weight × Sensitivity Score," the comprehensive score for each product category production line is calculated. Finally, the product categories are ranked from highest to lowest according to their comprehensive scores, determining the product category priority as "Coconut Water Production Line > Raw Coconut Milk Production Line > Frozen Coconut Cream Production Line," and a priority list is constructed accordingly. The core function of the priority list is: when conflicts cannot be resolved, the non-core batches of lower-priority production lines can be de-loaded or production delayed to ensure the operation of higher-priority production lines.

[0095] S27. Combine and check for conflicts by adjusting the adjustable time amount, execution sequence list and priority list to generate candidate scheduling instruction schemes.

[0096] In this embodiment of the invention, the adjustable time amount, execution sequence list, and priority list are combined according to the "conflict mitigation maximization" strategy to generate 2-3 sets of candidate schemes (e.g., scheme 1: only implement timing adjustment; scheme 2: timing adjustment combined with low priority batch delay). Triple conflict verification is performed on the candidate schemes, specifically including: (1) time conflict verification: check whether there is overlap in the start time of each process after adjustment; (2) process constraint verification: confirm that the material waiting time does not exceed the maximum allowable waiting shelf life (e.g., the waiting time of coconut water raw materials is ≤4 hours); (3) production target verification: ensure that the delivery of key orders is not delayed. If a conflict is found in the verification (e.g., a certain scheme causes the waiting time of coconut water raw materials to reach 5 hours, exceeding the maximum allowable waiting shelf life), then return to S24 to recalculate the adjustable time amount, iterate and optimize until a conflict-free candidate scheme is generated, thereby obtaining the scheduling instruction candidate scheme.

[0097] S28. Evaluate the effectiveness of candidate scheduling instructions in mitigating load conflicts and the disturbances to the production plan. Select the optimal solution based on the production optimization objective and format it as an active scheduling instruction output.

[0098] In this embodiment of the invention, the active scheduling instruction includes an optimization scheme for the start-up sequence of high-energy-consuming processes in each production line. The evaluation index includes two core dimensions: (1) conflict mitigation effect (such as the percentage reduction of peak load and the reduction of over-constraint duration), where the percentage reduction of peak load = (peak load of conflict period before adjustment - peak load of conflict period after adjustment) ÷ peak load of conflict period before adjustment × 100%; the reduction of over-constraint duration = (over-constraint duration before adjustment - over-constraint duration after adjustment) ÷ over-constraint duration before adjustment × 100%, and the data are all from the peak data and over-constraint period statistics before and after the adjustment of the total energy demand load curve; (2) production disturbance degree (such as order delay duration and batch adjustment quantity), where the order delay duration = adjusted order delivery time - original planned delivery time (the original planned delivery time comes from the production plan data, and the adjusted delivery time is calculated by reversing the adjusted process duration); the batch adjustment quantity = the number of batches with start-up time adjustments, and the data comes from the batch adjustment records in the scheduling scheme.

[0099] The evaluation logic is as follows: Under the premise of meeting the production optimization goal (zero delays for key orders), select the solution with the optimal conflict mitigation effect. For example: After adjusting Solution 1, the peak load during the conflict period decreases from 1420kW (previously calculated by aggregation in step 103) to 1180kW, with a peak load reduction percentage of (1420-1180) ÷ 1420 × 100% ≈ 17%; the over-constraint duration is shortened from 40 minutes (previously calculated by statistics in step S22) to 0 minutes, a reduction of 100%; there are no order delays, and the number of batches adjusted is 2. After adjusting Solution 2, the peak load decreases to 1150kW, a reduction percentage of ≈19%; the over-constraint duration is 0 minutes; however, non-urgent orders for frozen coconut milk are delayed by 2 hours (originally scheduled delivery time 16:00, adjusted delivery time 18:00), and the number of batches adjusted is 3. Because Option 2 violates the "zero delay for key orders" target (although the delays are for non-key orders, key orders must be prioritized while minimizing delays for non-key orders), Option 1 is selected first. The final output of the active scheduling instruction adopts a standardized JSON format. The format fields and data types are determined by the industrial control protocol specification, including core information such as the production line identifier (string type, such as "YZ-01" representing coconut water production line 01), the name of the target high-energy-consuming process (string type, such as "evaporation and concentration process"), the adjusted start time (timestamp format), and the execution priority (numerical type, 1 is the highest priority).

[0100] Step 105: Based on the active scheduling instruction, generate and issue corresponding timing control instructions to the control systems of each production line to dynamically adjust the actual start-up time of high-energy-consuming processes in each production line, so that the real-time energy consumption load of the production system matches the optimization scheme.

[0101] Furthermore, step 105 may include the following sub-steps:

[0102] S31. Parse the active scheduling instruction, extract the adjustable time amount, execution sequence list and priority list associated with each production line, and allocate the adjustable time amount according to the production line identifier based on the execution sequence list and priority list to generate independent control tasks corresponding to each production line.

[0103] In this embodiment of the invention, the active scheduling instruction parsing process is as follows: the standardized JSON format active scheduling instruction output in step 104 is parsed through the timing control instruction generation module built into the edge computing server. The core of the parsing is to extract three types of key data, which are derived from the corresponding preceding steps: First, the adjustable time amount, derived from the calculation result of step S24, includes the delay or early start time of each target high-energy-consuming process, such as the 20-minute delay of the evaporation and concentration process of the YZ-01 production line; second, the execution sequence list, derived from the construction result of step S25, includes the staggered start order of similar high-energy-consuming processes, such as the order of YZ-01 evaporation and concentration starting at 10:00 and SY-01 evaporation and concentration starting at 10:10; third, the priority order list, derived from the construction result of step S26, specifically ordered as coconut water production line above raw coconut milk production line, and raw coconut milk production line above frozen coconut milk production line. During the parsing process, data integrity is ensured through JSON field validation. Validation fields include production line identifier, process name, adjustable duration, sequence order, etc. If any field is missing or has an incorrect format, a parsing exception signal will be generated and fed back to the active scheduling instruction generation module.

[0104] Independent control task generation and allocation process: Based on the three types of data extracted through parsing, data allocation is performed with the production line identifier as the core dimension. The allocation logic is as follows: according to the association between production lines and processes in the execution sequence table, the corresponding adjustable time amount is bound to the specific production line. At the same time, combined with the priority list, high-priority production lines (such as coconut water production lines) are marked with priority execution identifiers, and degraded batches of low-priority production lines (such as frozen coconut milk production lines) are marked with load reduction execution identifiers. The final generated independent control tasks are structured data. Each task contains core fields such as production line identifier, target high-energy-consuming process name, adjustable time amount, execution sequence number, and execution priority. A typical example of a production line identifier is YZ-01, an typical example of an adjustable time amount is a 20-minute delay, and a typical example of an execution priority is level 1.

[0105] S32. Verify whether the independent control tasks meet the preset constraints of the corresponding production line and generate verification data. The preset constraints include: the material waiting time caused by process adjustment does not exceed its maximum allowable shelf life, and the production line that has been included in the priority list for subsequent downgrading is in a state where it is safe to reduce load.

[0106] In this embodiment of the invention, the definition of the preset constraints strictly follows the characteristics of coconut product production, specifically including two core constraints: (1) Material waiting time constraint: the material waiting time after process adjustment does not exceed its maximum allowable shelf life; (2) Downgraded production line safety status constraint: the production line that is subsequently downgraded in the priority list (such as the frozen coconut milk non-urgent batch production line) must be in a state where it is safe to reduce load. The judgment criteria are that the current batch is a non-core batch, the equipment operating parameters after load reduction are within the rated range (such as compressor speed ≥300r / min, data from equipment factory parameters), and the subsequent process has a sufficient buffer period.

[0107] The process of generating verification data is as follows: The timing control instruction generation module calls the corresponding production line process parameters (maximum allowable shelf life), production plan data (start time of subsequent processes) and real-time equipment operation data (transmitted by the production line PLC control system through PROFIBUS-DP bus, sampling frequency 1 second / time, including equipment speed, current, etc.) stored in the local database of the edge computing server, and verifies each independent control task one by one: (1) Calculate the material waiting time after process adjustment = end time of adjusted process - end time of previous process (end time of adjusted process = original planned end time + adjustable time amount, the original planned end time comes from the production plan data in step 101), and compare it with the maximum allowable shelf life; (2) Check whether the downgraded production line meets the three judgment criteria for safe load reduction. The final generated verification data is a structured result. Each data entry includes "control task ID, production line identifier, verification constraint type, verification result (pass / fail), reason for failure (e.g., material waiting time 5 hours > 4 hours), and relevant data source identifier." The "relevant data source identifier" can be traced back to the specific process standard document number, production plan batch number, or equipment data acquisition timestamp, ensuring data verifiability and strengthening the support of the claims. If the verification fails, the verification data is fed back to step S24 to recalculate the adjustable time, forming a closed-loop optimization; if the verification passes, the verification data will serve as a necessary basis for the generation of subsequent instructions.

[0108] S33. When the verification data passes the verification, based on the adjustment time of the independent control task corresponding to the verification data and its sorting position in the execution sequence list, calculate the planned start timestamp of the high-energy-consuming process of the production line and generate a timing control instruction containing the planned start timestamp.

[0109] In this embodiment of the invention, after the verification data passes, the adjustable time amount and execution sequence number in the independent control task generated in step S31 are combined with the original planned start time in the production plan in step 101. The verification in step S32 confirms no constraint conflicts through data verification, thereby advancing the calculation of the plan start timestamp and the generation of timing control instructions. The plan start timestamp is calculated according to... The result is calculated by adding the adjustable time amount and the sequence offset duration, which is determined based on the sorting position in the execution sequence list. The offset duration for the process with sequence number 1 is set to 0 minutes for priority startup. The offset duration for sequence number n (n≥2) is (n-1) multiplied by the minimum peak-shifting interval. The minimum peak-shifting interval is set to 5 minutes, derived from experiments on the energy consumption fluctuation characteristics of the target high-energy-consuming processes. A 5-minute interval ensures that the peak load of the preceding process decreases by more than 30%, avoiding load superposition, and reducing the risk of load superposition by 40% compared to a 3-minute interval.

[0110] The original planned start time for the evaporation and concentration process of the coconut water production line YZ-01 was... The original planned start time T0 for the evaporation and concentration process of the raw coconut milk production line SY-01 was 9:40. With an adjustable time delay of 20 minutes and execution sequence number 1, the planned start timestamp is 9:40 plus 20 minutes plus 0 minutes, resulting in 10:00. The adjusted time delay of T0 is 25 minutes and execution sequence number 2, resulting in 9:45 plus 25 minutes plus 5 minutes, resulting in 10:15. The timestamp format uses the style 2026-XX-XX 10:00:00, maintaining consistency with the timestamp format of the baseline timeline mentioned earlier. Based on the calculated planned start timestamps, standardized timing control instructions are generated by combining the corresponding production line identifier and target high-energy-consuming process information, ensuring that the instructions can be directly parsed and executed by the production line control system.

[0111] S34. Send the timing control instructions corresponding to each production line to the corresponding production line control system, and monitor the actual start time of high-energy-consuming processes.

[0112] In this embodiment of the invention, the instruction issuance process is as follows: The timing control instruction generation module sends the timing control instructions corresponding to each production line to the production line PLC control system via an industrial communication network (using the previously mentioned combination of PROFIBUS-DP bus and TCP / IP protocol). The issuance employs a "point-to-point + broadcast confirmation" mechanism: First, the instruction is sent point-to-point to the target production line PLC control system, and simultaneously, the issuance information is broadcast to the monitoring module of the edge computing server. After receiving the instruction, the PLC control system sends back an "instruction reception confirmation" signal. If no confirmation signal is received within 1 second (consistent with the previously mentioned transmission delay threshold), a retransmission mechanism is triggered. The number of retransmissions does not exceed 3. If confirmation is still not received, an issuance anomaly alarm is generated and pushed to the production scheduling terminal. This issuance mechanism ensures the reliability of instruction transmission, avoids execution omissions due to communication interruptions, and strengthens the correspondence between the method and the production system communication module.

[0113] Actual startup time monitoring process: Through the PLC control system deployed on each target high-energy-consuming process equipment, the system collects equipment startup signals in real time (e.g., a successful startup is determined when the motor operating current is greater than 10% of the rated current), extracts the system time corresponding to the startup signal, and generates actual startup time data (the timestamp format is consistent with the planned startup timestamp). The monitoring module collects actual startup time data of each production line PLC control system at a frequency of 1 second / time, stores it in the local database of the edge computing server, and associates it in real time with the planned startup timestamp of the corresponding timing control instruction to form an "instruction-execution" data traceability chain, providing raw data for subsequent deviation judgment. The data collection frequency is determined based on the equipment startup response speed.

[0114] S35. When the deviation between the actual start time and the corresponding planned start timestamp exceeds a preset threshold, automatic fine-tuning of the cycle time of subsequent processes is triggered.

[0115] In this embodiment of the invention, the deviation determination process is as follows: calculate the time difference between the actual start time and the planned start timestamp (deviation value = actual start time - planned start timestamp), and preset the deviation threshold to ±2 minutes (this threshold is determined by two aspects: (1) equipment control accuracy: the time control accuracy of the production line PLC control system is ±1 minute, with a 1-minute redundancy reserved; (2) process influence: experimental verification shows that when the deviation is ≤2 minutes, the material waiting time is still within the maximum allowable shelf life and has no impact on product quality; when the deviation is >3 minutes, the risk of the coconut water material waiting time exceeding the standard increases to 30%). If the absolute value of the deviation is ≤2 minutes, it is determined to be "deviation qualified" and no adjustment is required; if the absolute value of the deviation is >2 minutes, it is determined to be "deviation exceeding the standard" and triggers the subsequent process cycle fine-tuning mechanism.

[0116] The generation and execution process of subsequent process cycle time fine-tuning data is as follows: The fine-tuning mechanism is based on the core principles of "ensuring that the material waiting time does not exceed the maximum allowable shelf life" and "not affecting the continuity of core production batches." The specific calculation logic for fine-tuning data is as follows: Based on the duration of deviation exceeding the standard, the adjustment amount is allocated according to the rule of "subsequent process cycle time compression / extension = deviation exceeding the standard duration ÷ number of subsequent processes." For example, if the coconut water production line's evaporation and concentration process is planned to start at 10:00, but actually starts at 10:03 (deviation +3 minutes, exceeding the standard by 1 minute), and the subsequent processes are freezing and packaging, a total of 2 steps, then the cycle time of each process will be compressed by 0.5 minutes. The process of generating fine-tuning data: The monitoring module extracts the information of subsequent processes from the production plan of the production line (derived from step 101), calculates the compressible / extended duration of each subsequent process (based on the shortest / longest allowable duration of the process in the process standard document, such as the minimum allowable duration of the freezing process being 25 minutes, originally planned to be 30 minutes, which can be compressed by 5 minutes), and generates fine-tuning instruction data containing "subsequent process name, fine-tuning direction (compression / extension), fine-tuning duration, and adjusted process time period", which is then sent to the corresponding process equipment through the PLC control system. After fine-tuning, the monitoring module continuously collects the actual operating data of the subsequent processes to verify whether the material waiting time has returned to the allowable range (e.g., after the adjustment of the coconut water freezing process, the material waiting time decreased from 4.2 hours to 3.8 hours, meeting the requirement of ≤4 hours), ensuring that the real-time energy consumption load of the production system returns to the optimized solution range.

[0117] Please see Figure 2 , Figure 2 This is a structural block diagram of a production line control system for coconut products provided in an embodiment of the present invention.

[0118] This invention provides a production line control system for coconut products, applicable to a production system comprising at least two independent production lines, characterized in that the system includes:

[0119] The production plan data acquisition module 201 is used to acquire production plan data issued by the manufacturing execution system.

[0120] The energy consumption prediction module 202 is used to input production plan data into a pre-trained energy consumption prediction model, which generates a predicted energy consumption sequence for each production line during the corresponding planned production period.

[0121] The load curve aggregation module 203 is used to aggregate all predicted energy consumption sequences and generate the total energy demand load curve of the production system in a preset future period.

[0122] The active scheduling instruction generation module 204 is used to feed back active scheduling instructions based on the total energy demand load curve and real-time energy supply status. The active scheduling instructions include optimization schemes for the start-up sequence of high energy-consuming processes in each production line.

[0123] The timing control instruction execution module 205 is used to generate and send corresponding timing control instructions to the control systems of each production line according to the active scheduling instructions, so as to dynamically adjust the actual start time of high energy-consuming processes in each production line and match the real-time energy consumption load of the production system with the optimization scheme.

[0124] The specific implementation method of the production line control system for coconut products is basically the same as the specific implementation method of the production line control method for coconut products described above, and will not be repeated here.

[0125] Please see Figure 3 , Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention.

[0126] An electronic device according to an embodiment of the present invention includes: a memory 301 and a processor 302. The memory 301 stores a computer program. When the computer program is executed by the processor 302, the processor 302 executes the production line control method for coconut products as described in any of the above embodiments.

[0127] Memory 301 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Memory 301 has storage space 303 for program code 313 for performing any of the method steps described above. For example, storage space 303 for program code may include various program codes 313 for implementing the various steps in the methods described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When run by a computing processing device, this code causes the computing processing device to perform the various steps in the methods described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When these codes are run by a computing processing device, the device causes it to execute the various steps in the production line control method for coconut products described above.

[0128] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a production line control method for coconut products as described in any of the above embodiments.

[0129] This invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer performs a production line control method for coconut products as described in any of the above embodiments.

[0130] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0131] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0132] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0133] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0134] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0135] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A production line control method for coconut products, applied to a production system comprising at least two independent production lines, characterized in that, The method includes: Obtain production plan data issued by the Manufacturing Execution System; The production plan data is input into a pre-trained energy consumption prediction model, which generates a predicted energy consumption sequence for each production line during the corresponding planned production period. Aggregate all the predicted energy consumption sequences to generate the total energy demand load curve of the production system within a preset future period; Based on the total energy demand load curve and real-time energy supply status, active scheduling instructions are fed back; the active scheduling instructions include an optimization scheme for the start-up sequence of high energy-consuming processes in each of the production lines. Based on the active scheduling instruction, corresponding timing control instructions are generated and sent to the control systems of each production line to dynamically adjust the actual start-up time of high-energy-consuming processes in each production line, so that the real-time energy consumption load of the production system matches the optimization scheme.

2. The production line control method for coconut products according to claim 1, characterized in that, The energy consumption prediction model is pre-trained based on the historical production data of the production system; the historical production data includes at least: time-series power data, raw material input and output data for each process under different product categories.

3. The production line control method for coconut products according to claim 1, characterized in that, The predicted energy consumption sequence includes an energy consumption sub-sequence targeting a high-energy-consuming process; the high-energy-consuming process includes at least one of an instant sterilization process, a high-temperature sterilization process, a homogenization process, and an evaporation and concentration process.

4. The production line control method for coconut products according to claim 3, characterized in that, The step of aggregating all the predicted energy consumption sequences to generate the total energy demand load curve of the production system within a preset future period includes: Based on a preset future period, an initial reference time axis with a uniform time interval is set; The predicted energy consumption sequence corresponding to each production line is mapped onto the initial reference time axis according to the corresponding planned production period to generate a target reference time axis; wherein, for the idle period outside the planned production period, the energy consumption value on the target reference time axis is filled with zero; The total system load prediction value corresponding to each time point on the target reference time axis is calculated by summing up all the predicted energy consumption values ​​at each time point. Connect the predicted total system load values ​​at all time points on the target reference time axis to generate the total energy demand load curve corresponding to the production system.

5. The production line control method for coconut products according to claim 1, characterized in that, The step of feeding back active dispatch instructions based on the total energy demand load curve and real-time energy supply status includes: Obtain externally inputted energy constraints and production optimization objectives; wherein, the energy constraints include the upper limit of contracted electricity capacity or total energy consumption limit within a future set period; the production optimization objectives include ensuring timely delivery of key orders; The total energy demand load curve is compared with the energy constraints to determine the load conflict period; Based on the traceable load composition information in the total energy demand load curve, identify one or more target high-energy-consuming processes that contribute the most to the load during the load conflict period. For each of the target high-energy-consuming processes, within the time flexibility window allowed by the process of the target high-energy-consuming process, calculate the adjustable amount of time that the target high-energy-consuming process can be delayed or started earlier. When multiple target high-energy-consuming processes are of the same type and their planned time periods overlap within the load conflict period, an execution sequence table corresponding to the production system is constructed on the premise of ensuring the continuity of core production batches of each production line. When a conflict cannot be completely resolved by timing adjustments alone, a priority list corresponding to the production system is constructed based on the preset product category priority; wherein, the preset product category priority is set as follows: coconut water production line is higher than raw coconut milk production line, and raw coconut milk production line is higher than frozen coconut milk production line. The adjustable time amount, the execution sequence list, and the priority list are combined and conflict checked to generate candidate scheduling instruction schemes. The effectiveness of the candidate scheduling instructions in mitigating load conflicts and the disturbances to the production plan are evaluated. Based on the production optimization objectives, the optimal solution is selected and formatted as an active scheduling instruction output.

6. The production line control method for coconut products according to claim 5, characterized in that, The step of generating and issuing corresponding timing control instructions to the control systems of each production line according to the active scheduling instruction includes: The active scheduling instruction is parsed to extract the adjustable time amount, execution sequence list and priority list associated with each production line. Based on the execution sequence list and priority list, the adjustable time amount is allocated according to the production line identifier to generate independent control tasks corresponding to each production line. Each independent control task is verified to meet the preset constraints of the corresponding production line, and verification data is generated. The preset constraints include: the material waiting time caused by process adjustment does not exceed its maximum allowable shelf life, and the production line that is subsequently downgraded in the priority list is in a state where it is safe to reduce load. When the verification data passes the verification, the planned start timestamp of the corresponding high-energy-consuming process on the production line is calculated based on the adjustment time of the independent control task corresponding to the verification data and its sorting position in the execution sequence list, and a timing control instruction containing the planned start timestamp is generated. The timing control commands corresponding to each production line are sent to the corresponding production line control system, and the actual start-up time of high-energy-consuming processes is monitored. When the deviation between the actual start time and the corresponding planned start timestamp exceeds a preset threshold, automatic fine-tuning of the cycle time of subsequent processes is triggered.

7. A production line control system for coconut products, applied to a production system comprising at least two independent production lines, characterized in that, The system includes: The production planning data acquisition module is used to acquire production planning data issued by the manufacturing execution system. The energy consumption prediction module is used to input the production plan data into a pre-trained energy consumption prediction model, and the energy consumption prediction model generates a predicted energy consumption sequence for each production line during the corresponding planned production period. The load curve aggregation module is used to aggregate all the predicted energy consumption sequences to generate the total energy demand load curve of the production system in a preset future period. An active scheduling instruction generation module is used to feed back active scheduling instructions based on the total energy demand load curve and real-time energy supply status; the active scheduling instructions include an optimization scheme for the start-up sequence of high energy-consuming processes in each of the production lines. The timing control instruction execution module is used to generate and issue corresponding timing control instructions to the control systems of each production line according to the active scheduling instructions, so as to dynamically adjust the actual start time of high energy-consuming processes in each production line, so as to match the real-time energy consumption load of the production system with the optimization scheme.

8. An electronic device, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor causes the processor to perform the steps of the production line control method for coconut products as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the production line control method for coconut products as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, wherein when the program instructions are executed by a computer, the computer performs the production line control method for coconut products as described in any one of claims 1-6.

Images