Control system for adjusting beer heat energy center cold and hot balance based on AI large model

Abstract

The invention discloses a control system for adjusting beer heat energy center cold and heat balance based on an AI large model. The system comprises a control end and a server, wherein the control end comprises a first data acquisition module, a second data acquisition module and a communication module; the server pre-stores an AI large model and comprises a data processing module, a parameter determination module and a flow optimization module. The AI large model establishes a dynamic heat balance model and a heat balance incidence matrix based on sample data in a first production period, and executes pre-adjustment and real-time adjustment according to real-time data in a second production period. The heat value data representation method considering the non-conduction loss is innovatively defined, and the heat balance control precision is improved through combined calculation of the reference heat conduction coefficient and the dynamic heat conduction coefficient. Based on the heat balance incidence matrix, when any terminal device is subjected to adjustment operation, the system can predict the heat balance influence quantity of the adjustment on other terminals and carry out pre-compensation adjustment.

Description

Technical Field

[0001] This invention relates to the field of technology, and more specifically to a control system for regulating the thermal balance of beer based on a large AI model. Background Technology

[0002] Due to process requirements, breweries establish thermal energy centers. High-temperature heat energy from wort precooling and secondary steam from the boiling kettle is recovered and stored in the thermal energy center for wort preheating. After wort preheating, the thermal energy center still has surplus heat, which many breweries use in the packaging workshop to heat bottle washers, pasteurizers, and CIP systems, reducing steam consumption and lowering costs.

[0003] In recent years, with the rise of high-temperature heat pumps, beer factories have adopted a large number of high-temperature heat pumps to recover waste heat from the production process and provide high-temperature heat storage in the thermal energy center, so that the thermal energy center has a sufficient heat source for heating the packaging workshop.

[0004] Compared to the previous stable steam heating, the control unit could achieve stable temperature control at the heat-consuming end by setting fixed PID parameters. However, with the current use of thermal energy centers for heating, the flow rate and temperature at the heating end are constantly changing. Fixed PID parameters cannot stabilize the temperature at the heat-consuming end within the control range, resulting in large fluctuations, heat loss, and even affecting the equipment's quality control objectives.

[0005] By importing parameters such as temperature and flow rate at the heating end and the heat consumption end into the AI ​​model, the system automatically finds and matches the optimal PID parameters, thereby accurately controlling the target temperature at the heat consumption end within the set range and reducing steam consumption.

[0006] Meanwhile, the heating temperature requirements in the packaging workshop vary. The bottle washing machine requires a heating temperature of 90℃, the CIP machine requires 85℃, and the sterilizer requires 75℃. To minimize the return water temperature for better absorption of heat from the secondary steam in the boiling kettle by the heat energy center, a series heating method is used: first heating the bottle washing machine, then the CIP machine, and finally the sterilizer. However, this heating method involves complex multi-coupling control, requiring both sufficient heat to the heat-consuming areas and minimizing the return water temperature. This is difficult to achieve with current PLC control.

[0007] By importing relevant data into a large AI model, the system automatically finds and matches the optimal heating flow rate to achieve the above requirements. Summary of the Invention

[0008] To address the aforementioned problems, this invention provides a control system based on an AI large-scale model to regulate the thermal balance of the beer's heat energy center.

[0009] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0010] A control system based on an AI-powered large-scale model to regulate the thermal balance of beer brewing centers includes:

[0011] The control terminal includes a first data acquisition module, a second data acquisition module, and a communication module;

[0012] The first data acquisition module is configured to acquire calorific value data of the thermal energy center and each of the terminal devices;

[0013] The second data acquisition module is configured to acquire first parameter data and second parameter data of the thermal energy center and each of the terminal devices, wherein the first parameter data is a preset performance parameter and the second parameter data is a dynamic performance parameter;

[0014] The communication module is configured to transmit the calorific value data, the first parameter data, and the second parameter data to the server.

[0015] The server pre-stores large AI models, and the server includes a data processing module and a parameter determination module.

[0016] The data processing module is configured to receive the calorific value data, the first parameter data, and the second parameter data. The data processing module is used in the first production cycle, wherein the data of the first production cycle is provided as sample data to the AI ​​big model for training. The AI ​​big model establishes a dynamic heat balance model and a heat balance correlation matrix based on the sample data.

[0017] The parameter determination module is configured to receive real-time data collected in the second production cycle, and execute pre-adjustment instructions through the AI ​​big model based on the change between the real-time data and the ideal value, so that the calorific value data of each terminal device approaches the ideal value.

[0018] The flow optimization module is configured to perform real-time analysis of the dynamically changing data collected in real time during the second production cycle using the AI ​​big data model, and output adjustment commands to the regulating valves of each branch to keep the real-time sampled calorific value data within the allowable range of the ideal value.

[0019] The AI ​​big model, based on the thermal balance correlation matrix, predicts the thermal balance impact on other terminal devices when an adjustment operation is performed on the regulating valve of any terminal device, and performs pre-compensation adjustment through the flow optimization module to maintain the thermal balance of the entire system.

[0020] During the first production cycle, the AI ​​model determines the initial value of the heat balance correlation matrix based on the first parameter data and the second parameter data in the sample data through multivariate correlation analysis. During the second production cycle, when the second parameter data changes, the influence coefficient in the heat balance correlation matrix is ​​dynamically updated based on the similarity matching between the change and historical operating conditions.

[0021] In a preferred manner, the calorific value data is represented as a baseline thermal conductivity coefficient determined by the sampled terminal device under the first parameter data, and calculated by combining the dynamic thermal conductivity coefficient of the contents of the terminal device under the second parameter data;

[0022] The calorific value data includes first calorific value data and second calorific value data. The first calorific value data is the calorific value data reaching the terminal device, and the second calorific value data is the calorific value data when the terminal device reaches the ideal value.

[0023] The calorific value data is calculated based on temperature data and flow rate data, representing the total calorific value passing through the terminal device within a preset time period. The non-conductive loss is determined by the device structure parameters in the first parameter data and the environmental parameters in the second parameter data. The non-conductive loss is subtracted from the original calorific value data by the AI ​​big model to obtain the actual effective calorific value data for system control.

[0024] As a preferred approach, the AI ​​big model establishes the heat balance correlation matrix based on the first parameter data and the second parameter data in the sample data during the first production cycle. The heat balance correlation matrix represents the influence coefficient on the heat value data of other terminal devices when the heat value data of any terminal device changes.

[0025] When the second parameter data changes during the second production cycle, the AI ​​big model matches the amount of change with similar operating conditions in the historical operating condition database to determine the adjustment ratio of the change to each influence coefficient in the heat balance correlation matrix, and dynamically updates the heat balance correlation matrix based on the adjustment ratio to predict the amount of impact of the change on the heat balance of the entire system.

[0026] As a preferred approach, the dynamic heat balance model is constructed based on the heating network topology, pipeline heat conduction characteristics, and historical heat balance data. The AI ​​big model is trained using the sample data from the first production cycle to learn the dynamic characteristics of heat transfer between the heat energy center and each of the terminal devices.

[0027] During the second production cycle, the dynamic heat balance model predicts the heat source fluctuation trend of the thermal energy center based on the real-time collected second parameter data and the heat balance correlation matrix, and provides pre-adjustment parameters for the parameter determination module, so that the calorific value data of each terminal device approaches the ideal value before the heat source fluctuation occurs.

[0028] As a preferred approach, the heat balance correlation matrix is ​​obtained by performing multivariate correlation analysis on the changes in calorific value data of each terminal device during the first production cycle using the AI ​​large model. Each element in the heat balance correlation matrix represents the degree of influence on the calorific value data of other terminal devices when the calorific value data of a certain terminal device changes by one unit.

[0029] During the second production cycle, the AI ​​model determines the adjustment coefficients of each element in the thermal balance correlation matrix based on the real-time changes of the second parameter data through similar working conditions matching, and dynamically updates the element values ​​in the thermal balance correlation matrix based on the adjustment coefficients to reflect the thermal balance relationship under the current working conditions.

[0030] As a preferred embodiment, when the parameter determination module detects that the deviation between the calorific value data of any terminal device and the ideal value exceeds a first threshold, it calculates the pre-adjustment parameters through the AI ​​big model and sends the pre-adjustment command to the corresponding regulating valve.

[0031] The flow optimization module simultaneously calculates the impact of the pre-adjustment on the thermal balance of other terminal devices based on the thermal balance correlation matrix, and performs pre-compensation adjustment on the regulating valves of other branches according to the impact, so that the calorific value data of each terminal device remains within the allowable range of the ideal value during the pre-adjustment process, avoiding overall system imbalance caused by the adjustment of a single terminal device.

[0032] As a preferred approach, the flow optimization module defines an objective function, which includes minimizing the return water temperature of the thermal energy center and the pumping energy consumption. When the calorific value data of any terminal device is detected to be close to the boundary of its allowable range, the AI ​​big model calculates the impact of the flow adjustment of each branch on the calorific value data of other terminal devices based on the heat balance correlation matrix, and determines the calorific value requirements of the terminal devices according to priority.

[0033] By dynamically adjusting the opening ratio of the regulating valves in each branch, the return water temperature of the heat energy center is maintained within the target temperature range while meeting the calorific value requirements of each terminal device.

[0034] As a preferred approach, when a sudden operating condition is detected that causes a large fluctuation in the heat source, the parameter determination module quickly calculates the emergency adjustment parameters through the AI ​​large model. At the same time, the flow optimization module predicts the impact of the sudden operating condition on each terminal device based on the heat balance correlation matrix and executes a multi-level control strategy: first, adjust the flow distribution to quickly balance the system; second, adjust the control parameters to stabilize the temperature; and finally, adjust the heat source output of the thermal energy center to adapt to long-term operating condition changes.

[0035] The execution order and intensity of the multi-level control strategy are determined based on the magnitude of each influence coefficient in the thermal balance correlation matrix.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] This invention constructs a thermal balance correlation matrix using a large AI model, enabling accurate prediction of the impact of a single adjustment operation. It effectively solves the technical problem of system imbalance caused by local adjustment in traditional control systems, and significantly improves the stability of the thermal energy system in beer production.

[0038] This invention innovatively adopts a phased control strategy of the first and second production cycles. By using an AI big data model to establish an accurate dynamic thermal balance model in the first production cycle, and to achieve predictive control based on historical data in the second production cycle, the system’s adaptability and response speed to heat source fluctuations are greatly improved.

[0039] This invention defines a method for representing calorific value data that considers non-conductive losses. By combining the baseline thermal conductivity coefficient and the dynamic thermal conductivity coefficient, the actual effective calorific value data is accurately calculated, so that the calorific value data of each terminal device can be stably maintained within the allowable range of the ideal value, thereby improving the temperature control accuracy.

[0040] This invention implements a pre-compensation adjustment mechanism based on the thermal balance correlation matrix. When any terminal device is adjusted, the system can predict the impact of the adjustment on other terminals and perform pre-compensation, effectively avoiding system oscillations caused by a single adjustment and ensuring the continuity of the beer production process and product quality.

[0041] This invention employs a multi-level control strategy based on the magnitude of the influence coefficient, which can quickly restore system balance in the event of sudden operating conditions, prioritize the adjustment of terminal equipment that has a greater impact on the system, significantly reduce energy waste, improve the overall energy utilization efficiency of the thermal energy center, and achieve the production goal of energy conservation and consumption reduction. Attached Figure Description

[0042] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0043] Figure 1 A structural block diagram of the system provided in the embodiments of this disclosure. Detailed Implementation

[0044] 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, and 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.

[0045] This disclosure provides a control system based on an AI large-scale model to regulate the thermal balance of beer's core heat energy, such as... Figure 1 As shown, the system includes a control terminal and a server. The system aims to collect relevant data from the thermal energy center and various terminal devices through the control terminal, and transmit this data to the server for analysis and processing. This enables precise control and dynamic balance adjustment of the thermal energy system during beer production. In this embodiment, a brewery thermal energy center is used as the application scenario. However, a brewery thermal energy center is not the only applicable scenario for achieving the objectives of this invention; those skilled in the art can also apply this system to other industrial thermal energy systems according to actual conditions.

[0046] Specifically, the control terminal in this embodiment of the disclosure includes:

[0047] The first data acquisition module is configured to acquire calorific value data from the thermal energy center and each of the terminal devices. In this embodiment, the first data acquisition module uses a bottle washer, a CIP cleaning system, and a sterilizer as the main terminal devices, acquiring temperature and flow data at the outlet of the thermal energy center's heating end and at the inlet and outlet of each terminal device as calorific value data. Calorific value data reflects the distribution and flow of heat in the system, providing fundamental data support for subsequent heat balance control. The first data acquisition module monitors the calorific value data of each key node in real time using high-precision temperature sensors and flow meters, with a sampling frequency of once every 5 seconds to ensure the real-time performance and accuracy of the data.

[0048] The second data acquisition module is configured to acquire first parameter data and second parameter data from the thermal energy center and each of the terminal devices. The first parameter data consists of preset performance parameters, while the second parameter data consists of dynamic performance parameters. In this embodiment, the first parameter data includes inherent parameters of the equipment such as pipe diameter, wall thickness, and insulation material type; the second parameter data includes dynamically changing parameters such as ambient temperature, pipe surface temperature, and equipment operating status. The second data acquisition module collects these parameters through a distributed sensor network. The first parameter data is entered once during system installation, while the second parameter data is updated in real time, providing comprehensive system status information for the AI ​​large-scale model.

[0049] The first communication module is configured to transmit thermal data from the first data acquisition module and first parameter data and second parameter data from the second data acquisition module to the server. In this embodiment, the first communication module uses a combination of industrial Ethernet and wireless communication to transmit the acquired data to the server through an encrypted channel. The first communication module ensures the reliability and security of data transmission, providing timely and accurate data support to the server.

[0050] Specifically, the server in this embodiment of the disclosure includes:

[0051] The second communication module is configured to communicate with the first communication module and receive the calorific value data, the first parameter data, and the second parameter data. In this embodiment, the second communication module receives real-time data streams from the control terminal and performs preliminary data cleaning and formatting. The second communication module acts as a bridge for data interaction between the server and the control terminal, ensuring that the server can obtain complete and accurate system operation data.

[0052] The data processing module is configured to receive the calorific value data, the first parameter data, and the second parameter data. The data processing module distinguishes between the first production cycle and the second production cycle, using the data from the first production cycle as sample data to train the AI ​​large-scale model. The AI ​​large-scale model establishes a dynamic thermal balance model and a thermal balance correlation matrix based on the sample data. In this embodiment, the data processing module collects a large amount of sample data during the first 7 days of system operation (the first production cycle), including system operation data from different production batches and under different environmental conditions. The data processing module preprocesses this sample data, including outlier removal, data normalization, and feature extraction, providing high-quality input data for the training of the AI ​​large-scale model.

[0053] The parameter determination module is configured to receive real-time data collected during the second production cycle. Based on the change between the real-time data and the ideal value, it executes pre-adjustment instructions through the AI ​​model to bring the calorific value data of each terminal device closer to the ideal value. In this embodiment, the parameter determination module monitors the calorific value data of each terminal device in real time during the formal operation phase of the system (second production cycle) and compares it with the preset ideal value. When a deviation is detected, the parameter determination module calls the AI ​​model to calculate the optimal pre-adjustment parameters and sends instructions to the corresponding regulating valves to adjust the system state in advance and avoid large temperature fluctuations.

[0054] The flow optimization module is configured to perform real-time analysis of dynamically changing data collected in real time during the second production cycle using the AI ​​model, and output adjustment commands to the regulating valves of each branch to maintain the real-time sampled calorific value data within the allowable range of the ideal value. In this embodiment, the flow optimization module collects real-time system data every 15 seconds, performs rapid analysis using the AI ​​model, calculates the optimal opening degree of each branch regulating valve, and sends adjustment commands. The flow optimization module ensures that the system maintains thermal balance under various operating conditions, improving energy utilization efficiency.

[0055] Specifically, in the embodiments of this disclosure, the following is true:

[0056] The calorific value data is represented as the baseline thermal conductivity coefficient of the sampled terminal device under the first parameter data, calculated in conjunction with the dynamic thermal conductivity coefficient of the contents of the terminal device under the second parameter data. In this embodiment, for the bottle washing machine terminal, the baseline thermal conductivity coefficient is calculated based on the pipe material, diameter, and wall thickness; the dynamic thermal conductivity coefficient considers factors such as the beer bottle material, cleaning fluid composition, and temperature changes. The calorific value data includes first calorific value data and second calorific value data. The first calorific value data is the calorific value data reaching the terminal device, and the second calorific value data is the calorific value data when the terminal device reaches its ideal value. For example, for a bottle washing machine, the first calorific value data represents the calorific value of the hot water entering the bottle washing machine, and the second calorific value data represents the calorific value required for the bottle washing machine to reach its operating temperature of 90°C. The calorific value data is calculated based on temperature data and flow rate data, representing the total calorific value passing through the terminal device within a preset time period, and the non-conductive loss is jointly determined by the equipment structural parameters in the first parameter data and the environmental parameters in the second parameter data. In this embodiment, non-conductive losses include factors such as heat dissipation from pipes and heat dissipation from equipment surfaces. The AI ​​big data model learns these loss patterns through historical data and subtracts non-conductive losses from the original calorific value data to obtain actual effective calorific value data for system control.

[0057] Within the first production cycle, the AI ​​big data model establishes a heat balance correlation matrix based on the first and second parameter data from the sample data. This heat balance correlation matrix represents the influence coefficient of a change in the calorific value data of any terminal device on the calorific value data of other terminal devices. In this embodiment, the AI ​​big data model learns the heat balance relationship between the bottle washer, CIP cleaning system, and sterilizer by analyzing the sample data collected within the first production cycle. For example, when the water consumption of the bottle washer increases, it leads to a decrease in the inlet temperature of the sterilizer. The AI ​​big data model determines the specific degree of this influence through multivariate correlation analysis and establishes a heat balance correlation matrix. When the second parameter data changes within the second production cycle, the AI ​​big data model matches the change with similar operating conditions in the historical operating condition database to determine the adjustment ratio of each influence coefficient in the heat balance correlation matrix, and dynamically updates the heat balance correlation matrix based on the adjustment ratio. For example, when the ambient temperature drops by 5°C, the AI ​​big data model finds similar temperature change records in the historical operating condition database, determines the adjustment ratio of each influence coefficient in the heat balance correlation matrix, and enables the system to adapt to the new environmental conditions.

[0058] The dynamic heat balance model is constructed based on the heating network topology, pipeline heat conduction characteristics, and historical heat balance data. The AI ​​big data model is trained using the sample data from the first production cycle to learn the dynamic characteristics of heat transfer between the heat energy center and each terminal device. In this embodiment, the AI ​​big data model uses a deep learning algorithm to extract the heat transfer patterns between the heat energy center and each terminal device from the sample data of the first production cycle, including characteristics such as thermal inertia and response delay. In the second production cycle, the dynamic heat balance model predicts the heat source fluctuation trend of the heat energy center based on the real-time collected second parameter data and the heat balance correlation matrix, and provides pre-adjustment parameters for the parameter determination module. For example, when an increase in the heat load of the wort precooling system is detected, the dynamic heat balance model predicts that the heat source of the heat energy center will fluctuate and adjusts the parameters of each terminal device in advance to keep the system stable.

[0059] The heat balance correlation matrix is ​​obtained by performing multivariate correlation analysis on the changes in calorific value data of each terminal device during the first production cycle using the AI ​​large-scale model. Each element in the heat balance correlation matrix represents the degree of influence on the calorific value data of other terminal devices when the calorific value data of one terminal device changes by one unit. In this embodiment, the heat balance correlation matrix is ​​a 3×3 matrix (corresponding to the three terminal devices: bottle washer, CIP cleaning system, and sterilizer). The element a_ij in the matrix represents the degree of influence on the calorific value data of the i-th terminal device when the calorific value data of the j-th terminal device changes by one unit. During the second production cycle, the AI ​​large-scale model determines the adjustment coefficient of each element in the heat balance correlation matrix based on the real-time changes in the second parameter data through similar operating condition matching, and dynamically updates the element values ​​in the heat balance correlation matrix based on the adjustment coefficients. For example, when the CIP cleaning system switches its operating mode from standard cleaning to deep cleaning, the AI ​​large-scale model recognizes this change, adjusts the values ​​of relevant elements in the heat balance correlation matrix, and accurately reflects the new system state.

[0060] When the parameter determination module detects that the deviation between the calorific value data of any terminal device and the ideal value exceeds a first threshold, it calculates pre-adjustment parameters using the AI ​​large-scale model and sends pre-adjustment instructions to the corresponding regulating valves. In this embodiment, when the inlet temperature of the bottle washer is below 90°C and the deviation exceeds 2°C, the parameter determination module calls the AI ​​large-scale model to calculate the optimal pre-adjustment parameters, including the adjustment amount of the opening of the regulating valve at the heating end outlet and the pre-compensation amount of the regulating valve in the sterilizer branch. The flow optimization module simultaneously calculates the impact of the pre-adjustment on the heat balance of other terminal devices based on the heat balance correlation matrix, and performs pre-compensation adjustment on the regulating valves of other branches according to the impact amount. For example, when the flow rate of the bottle washer branch is increased, the flow optimization module predicts that this will cause the inlet temperature of the sterilizer to drop by 1.5°C, and adjusts the opening of the regulating valve in the sterilizer branch in advance to keep the calorific value data of each terminal device within the allowable range of the ideal value during the pre-adjustment process.

[0061] The flow optimization module defines an objective function, which includes minimizing the return water temperature of the thermal energy center and pumping energy consumption. When the calorific value data of any terminal device is detected to be close to the boundary of its allowable range, the AI ​​large model calculates the impact of flow adjustment of each branch on the calorific value data of other terminal devices based on the heat balance correlation matrix, and determines the calorific value requirements of the terminal devices according to priority. In this embodiment, the bottle washer is set as the highest priority, followed by the CIP cleaning system, and the sterilizer is the lowest priority. When system resources are limited, the calorific value requirements of the bottle washer are prioritized. By dynamically adjusting the opening ratio of the regulating valves of each branch, the return water temperature of the thermal energy center is maintained within the target temperature range while meeting the calorific value requirements of each terminal device. For example, when the bottle washer and sterilizer need to increase their flow rates simultaneously, the flow optimization module calculates the optimal allocation ratio to minimize the return water temperature while ensuring the temperature of the bottle washer.

[0062] When a sudden operating condition causes a significant fluctuation in the heat source, the parameter determination module quickly calculates emergency adjustment parameters using the AI ​​large-scale model. Simultaneously, the flow optimization module predicts the impact of the sudden operating condition on each terminal device based on the heat balance correlation matrix and executes a multi-level control strategy: first, adjusting flow distribution to quickly balance the system; second, adjusting control parameters to stabilize the temperature; and finally, adjusting the heat source output of the thermal energy center to adapt to long-term operating condition changes. In this embodiment, when a sudden boiler shutdown results in a 30% reduction in heat source, the system first adjusts the flow distribution of each branch to prioritize critical equipment; then, it adjusts the PID control parameters to stabilize the temperature; and finally, it adjusts the backup heat source as needed. The execution order and intensity of the multi-level control strategy are determined based on the magnitude of each influence coefficient in the heat balance correlation matrix. For example, for a bottle washer with a large influence coefficient on the return water temperature, the system will prioritize adjusting its flow distribution.

[0063] The control system for adjusting the thermal balance of the beer heat energy center based on an AI large model also includes a thermal balance monitoring module. This module is configured to receive thermal balance status data from the data processing module. When the return water temperature of the heat energy center exceeds a second threshold, it triggers an optimization command based on the thermal balance correlation matrix. In this embodiment, the thermal balance monitoring module monitors the return water temperature in real time, triggering an optimization command when it exceeds 70°C. The optimization command includes prioritizing the adjustment of flow distribution for terminal devices with larger influence coefficients on the return water temperature, and then adjusting the control parameters of terminal devices with smaller influence coefficients. For example, if the influence coefficient of the bottle washer on the return water temperature is 0.6, the CIP cleaning system is 0.3, and the sterilizer is 0.1, then the system prioritizes adjusting the flow distribution of the bottle washer to minimize the impact of the adjustment operation on the overall stability of the system.

[0064] This disclosure also provides a control method for adjusting the thermal energy center heat balance of beer based on an AI large model, including the following steps:

[0065] Step S1: In the first production cycle, the calorific value data of the thermal energy center and each terminal device is acquired through the first data acquisition module, and the first parameter data and the second parameter data are acquired through the second data acquisition module and transmitted to the server. In this embodiment, the system collects sample data during the first 7 days of initial operation, including system operation data under different production batches and different environmental conditions.

[0066] Step S2: The data processing module preprocesses the sample data and provides it to the AI ​​large model for training to establish a dynamic thermal balance model and a thermal balance correlation matrix. In this embodiment, the data processing module performs outlier removal, data normalization, and feature extraction, and the AI ​​large model establishes an accurate thermal balance model through deep learning algorithms.

[0067] Step S3: In the second production cycle, the parameter determination module receives real-time data. When the deviation between the calorific value data and the ideal value exceeds the first threshold, the module calculates the pre-adjustment parameters through the AI ​​big model and sends the pre-adjustment command. In this embodiment, the parameter determination module checks the system status every 15 seconds and performs pre-adjustment in a timely manner when a deviation is detected.

[0068] Step S4: The flow optimization module calculates the impact of pre-adjustment on the thermal balance of other terminal devices based on the thermal balance correlation matrix, and performs pre-compensation adjustment; in this embodiment, the flow optimization module predicts the impact of a single adjustment on other parts of the system and performs compensation in advance to avoid system oscillation.

[0069] Step S5: When a sudden operating condition is detected, a multi-level control strategy is executed: first, the flow distribution is adjusted; second, the control parameters are adjusted; and finally, the heat source output is adjusted. In this embodiment, the system determines the control sequence and intensity based on the magnitude of each influence coefficient in the heat balance correlation matrix.

[0070] Step S6: The cold and heat balance monitoring module monitors the return water temperature. When it exceeds the second threshold, it triggers an optimization command based on the heat balance correlation matrix. In this embodiment, when the return water temperature exceeds 70°C, the system automatically optimizes the flow distribution of each terminal device to reduce the return water temperature.

[0071] Step S7: The system continuously collects operational data, and the AI ​​big model updates the thermal balance correlation matrix regularly to improve the system control accuracy. In this embodiment, the AI ​​big model fine-tunes the thermal balance correlation matrix every 24 hours based on the latest data to adapt to long-term changes in the system.

[0072] This embodiment of the disclosure acquires calorific value data of the thermal energy center and each terminal device through a first data acquisition module, and acquires first parameter data and second parameter data through a second data acquisition module, providing the server with comprehensive system status information. This data acquisition method ensures that the system can accurately grasp the operating status of the thermal energy center, laying the foundation for precise control.

[0073] This embodiment of the disclosure distinguishes between the first and second production cycles through a data processing module. It uses sample data from the first production cycle to train a large AI model, establishing a dynamic thermal balance model and a thermal balance correlation matrix. In the second production cycle, it implements predictive control based on historical data. This phased control strategy significantly improves the system's adaptability and response speed to heat source fluctuations, making temperature control in the beer production process more stable and reliable.

[0074] This disclosure defines a method for representing calorific value data that considers non-conductive losses. By combining a baseline thermal conductivity coefficient and a dynamic thermal conductivity coefficient, the actual effective calorific value data is accurately calculated. This method enables the system to more accurately reflect the true situation of heat energy transfer, improves the accuracy of temperature control, and ensures that the calorific value data of each terminal device can be stably maintained within the allowable range of ideal values.

[0075] This disclosure implements a pre-compensation adjustment mechanism based on a thermal balance correlation matrix. When any terminal device is adjusted, the system can predict the impact of the adjustment on other terminals and perform pre-compensation. This mechanism effectively avoids system oscillations caused by a single adjustment, ensuring the continuity of the beer production process and product quality, and shows significant advantages, especially in complex systems where multiple terminal devices work together.

[0076] This disclosed embodiment employs a multi-level control strategy based on the magnitude of the influence coefficient, enabling rapid restoration of system balance in the event of sudden operational disruptions. The system prioritizes adjustments to terminal equipment with the greatest impact on the system. This targeted adjustment method significantly reduces energy waste, improves the overall energy utilization efficiency of the thermal energy center, achieves energy-saving and consumption-reducing production goals, and simultaneously ensures the process quality of beer production.

[0077] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, function, and operation of possible implementations of apparatus, methods, and computer program products according to embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than those disclosed in the description; sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based device that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A control system for regulating the thermal balance of beer brewing center based on an AI large-scale model, characterized in that, The control terminal includes a first data acquisition module, a second data acquisition module, and a communication module; The first data acquisition module is configured to acquire calorific value data of the thermal energy center and each of the terminal devices; The second data acquisition module is configured to acquire first parameter data and second parameter data of the thermal energy center and each of the terminal devices, wherein the first parameter data is a preset performance parameter and the second parameter data is a dynamic performance parameter; The communication module is configured to transmit the calorific value data, the first parameter data, and the second parameter data to the server. The server pre-stores large AI models, and the server includes a data processing module and a parameter determination module. The data processing module is configured to receive the calorific value data, the first parameter data, and the second parameter data. The data processing module is used in the first production cycle, wherein the data of the first production cycle is provided as sample data to the AI ​​big model for training. The AI ​​big model establishes a dynamic heat balance model and a heat balance correlation matrix based on the sample data. The parameter determination module is configured to receive real-time data collected in the second production cycle, and execute pre-adjustment instructions through the AI ​​big model based on the change between the real-time data and the ideal value, so that the calorific value data of each terminal device approaches the ideal value. The flow optimization module is configured to perform real-time analysis of the dynamically changing data collected in real time during the second production cycle using the AI ​​big data model, and output adjustment commands to the regulating valves of each branch to keep the real-time sampled calorific value data within the allowable range of the ideal value. The AI ​​big model, based on the thermal balance correlation matrix, predicts the thermal balance impact on other terminal devices when an adjustment operation is performed on the regulating valve of any terminal device, and performs pre-compensation adjustment through the flow optimization module to maintain the thermal balance of the entire system. During the first production cycle, the AI ​​model determines the initial value of the heat balance correlation matrix based on the first parameter data and the second parameter data in the sample data through multivariate correlation analysis. During the second production cycle, when the second parameter data changes, the influence coefficient in the heat balance correlation matrix is ​​dynamically updated based on the similarity matching between the change and historical operating conditions.

2. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 1, characterized in that, The calorific value data is represented as the baseline thermal conductivity coefficient determined by the sampled terminal device under the first parameter data, and calculated by combining the dynamic thermal conductivity coefficient of the contents of the terminal device under the second parameter data; The calorific value data includes first calorific value data and second calorific value data. The first calorific value data is the calorific value data reaching the terminal device, and the second calorific value data is the calorific value data when the terminal device reaches the ideal value. The calorific value data is calculated based on temperature data and flow rate data, representing the total calorific value passing through the terminal device within a preset time period. The non-conductive loss is determined by the device structure parameters in the first parameter data and the environmental parameters in the second parameter data. The non-conductive loss is subtracted from the original calorific value data by the AI ​​big model to obtain the actual effective calorific value data for system control.

3. The control system for adjusting the thermal balance of beer based on an AI large model according to claim 2, characterized in that, During the first production cycle, the AI ​​big model establishes the heat balance correlation matrix based on the first parameter data and the second parameter data in the sample data. The heat balance correlation matrix represents the influence coefficient on the heat value data of other terminal devices when the heat value data of any terminal device changes. When the second parameter data changes during the second production cycle, the AI ​​big model matches the amount of change with similar operating conditions in the historical operating condition database to determine the adjustment ratio of the change to each influence coefficient in the heat balance correlation matrix, and dynamically updates the heat balance correlation matrix based on the adjustment ratio to predict the amount of impact of the change on the heat balance of the entire system.

4. The control system for adjusting the thermal balance of beer based on an AI large model according to claim 3, characterized in that, The dynamic heat balance model is constructed based on the heating network topology, pipeline heat conduction characteristics and historical heat balance data. The AI ​​big model is trained by the sample data of the first production cycle to learn the dynamic characteristics of heat transfer between the heat energy center and each of the terminal devices. During the second production cycle, the dynamic heat balance model predicts the heat source fluctuation trend of the thermal energy center based on the real-time collected second parameter data and the heat balance correlation matrix, and provides pre-adjustment parameters for the parameter determination module, so that the calorific value data of each terminal device approaches the ideal value before the heat source fluctuation occurs.

5. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 4, characterized in that, The heat balance correlation matrix is ​​obtained by performing multivariate correlation analysis on the changes in calorific value data of each terminal device during the first production cycle using the AI ​​big model. Each element in the heat balance correlation matrix represents the degree of influence on the calorific value data of other terminal devices when the calorific value data of a certain terminal device changes by one unit. During the second production cycle, the AI ​​model determines the adjustment coefficients of each element in the thermal balance correlation matrix based on the real-time changes of the second parameter data through similar working conditions matching, and dynamically updates the element values ​​in the thermal balance correlation matrix based on the adjustment coefficients to reflect the thermal balance relationship under the current working conditions.

6. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 5, characterized in that, When the parameter determination module detects that the deviation between the calorific value data of any terminal device and the ideal value exceeds a first threshold, it calculates the pre-adjustment parameters through the AI ​​big model and sends the pre-adjustment command to the corresponding regulating valve. The flow optimization module simultaneously calculates the impact of the pre-adjustment on the thermal balance of other terminal devices based on the thermal balance correlation matrix, and performs pre-compensation adjustment on the regulating valves of other branches according to the impact, so that the calorific value data of each terminal device remains within the allowable range of the ideal value during the pre-adjustment process, avoiding overall system imbalance caused by the adjustment of a single terminal device.

7. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 6, characterized in that, The flow optimization module defines an objective function, which includes minimizing the return water temperature of the thermal energy center and the pumping energy consumption. When the calorific value data of any terminal device is detected to be close to the boundary of its allowable range, the AI ​​big model calculates the impact of the flow adjustment of each branch on the calorific value data of other terminal devices based on the heat balance correlation matrix, and determines the calorific value requirements of the terminal devices according to priority. By dynamically adjusting the opening ratio of the regulating valves in each branch, the return water temperature of the heat energy center is maintained within the target temperature range while meeting the calorific value requirements of each terminal device.

8. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 7, characterized in that, When a sudden operating condition is detected that causes a large fluctuation in the heat source, the parameter determination module quickly calculates the emergency adjustment parameters through the AI ​​big model. At the same time, the flow optimization module predicts the impact of the sudden operating condition on each terminal device based on the heat balance correlation matrix and executes a multi-level control strategy: first, adjust the flow distribution to quickly balance the system; second, adjust the control parameters to stabilize the temperature; and finally, adjust the heat source output of the thermal energy center to adapt to long-term operating condition changes. The execution order and intensity of the multi-level control strategy are determined based on the magnitude of each influence coefficient in the thermal balance correlation matrix.

9. The control system for adjusting the thermal balance of beer's heat energy center based on an AI large model according to claim 8, characterized in that, It also includes a heat balance monitoring module, which is configured to receive heat balance status data from the data processing module and trigger an optimization instruction based on the heat balance correlation matrix when the return water temperature of the heat energy center is detected to exceed the second threshold. The optimization instructions include prioritizing the flow distribution of terminal devices with larger influence coefficients on return water temperature, and then adjusting the control parameters of terminal devices with smaller influence coefficients, in order to minimize the impact of the adjustment operation on the overall stability of the system.

Images