Cold storage temperature prediction system and method based on mathematical modeling and fitting

Abstract

The invention discloses a refrigeration house temperature prediction system and method based on mathematical modeling and fitting, and belongs to the technical field of computer science. Comprising a sensor module, a data acquisition module, a data processing module and a prediction module, and the method comprises the steps of sensor deployment and data acquisition, data cleaning and preprocessing, data segmentation, model establishment and fitting, and prediction and evaluation. The method is simple in design and easy to integrate and maintain, and the fan operation strategy can be optimized and the energy efficiency can be improved by accurately predicting the cold room temperature change. And the method is high in adaptability and can adapt to different cold storage environments and changes of various influence factors. Fitting and a particle swarm optimization algorithm are adopted, and accurate real-time prediction of cold room temperature changes can be achieved. The problems that a traditional physical mechanism model is numerous in complex influence factors, complex in interaction and incapable of accurately modeling are effectively solved, the hysteresis quality of temperature changes is overcome, the prediction result is more timely and accurate, and therefore the reliability of refrigeration house temperature control is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of computer science and technology, and in particular to a cold storage temperature prediction system and method based on mathematical modeling and fitting. Background Art

[0002] Traditional cold storage management methods mainly rely on manual operation, which is not only inefficient but also prone to errors. This management method is highly dependent on personnel, resulting in increased labor costs, and the reaction speed and accuracy in temperature control cannot meet the requirements of modern cold chain logistics for efficient and precise temperature control. With the development of new technologies, cold storage management is gradually moving towards automation and intelligence. In particular, the energy-saving control of cold storage has gradually changed from manual operation to automated system management. In the energy-saving automation control of cold storage, the control of cold room temperature is a key step. This process relies on the monitoring and adjustment of the cold room temperature to ensure efficient use of energy and reduced operating costs while meeting storage requirements. Accurately predicting the change in cold room temperature under different fan-on conditions is the core issue of temperature control and a priority issue that needs to be solved in cold storage automation energy saving.

[0003] The change of cold room temperature is affected by many factors, including outdoor temperature, wind speed, cargo accumulation, door opening and closing frequency, and fan defrosting. These factors cause the temperature change to have hysteresis and nonlinear characteristics. Traditional temperature control methods are often based on empirical rules or simple linear models, which cannot accurately reflect the comprehensive impact of these complex factors on temperature, and thus cannot achieve accurate temperature control.

[0004] In the past, attempts were made to predict cold room temperature changes by building physical models based on differential equations. However, due to the complexity and variability of cold storage systems, these models are difficult to accurately describe the dynamic characteristics of temperature changes. Physical models require accurate modeling of each influencing factor, and the complex coupling between these factors makes the implementation of physical modeling extremely difficult, and the model complexity and computational cost are high, making it less practical.

[0005] Currently, more research tends to use data-driven methods for temperature prediction, such as multivariate time series analysis based on neural networks. Although this method can capture the periodic pattern of temperature changes to a certain extent, it relies on training with historical data and has limited adaptability to external changes. Especially when it is necessary to predict the future changes in cold room temperature under specific conditions (such as different fan opening conditions), the limitations of this method become apparent because it cannot flexibly simulate temperature changes under different operating conditions.

[0006] In addition, existing data-driven models are usually based on empirical and statistical methods, which also incorporate the start and stop patterns of fans into the model learning, which means that model predictions rely on past operating patterns rather than responding to new control strategies. This limitation makes it difficult for traditional empirical methods to accurately predict and control temperature when faced with complex multi-factor changes in cold storage systems. Summary of the invention

[0007] The purpose of the present invention is to provide a cold storage temperature prediction system and method based on mathematical modeling and fitting. By systematically analyzing and modeling various influencing factors, the change of cold room temperature can be predicted in real time, which is of great significance to the energy-saving automatic control of cold storage. It can not only improve the accuracy of temperature control, but also provide flexible temperature prediction and control strategies under different operating conditions, thereby achieving more efficient cold storage management.

[0008] To achieve the above-mentioned purpose, the present invention provides a cold storage temperature prediction system based on mathematical modeling and fitting, including a sensor module, a data acquisition module, a data processing module and a prediction module, the sensor module, the data acquisition module, the data processing module and the prediction module are connected in sequence, the sensor module is responsible for real-time monitoring of the environmental parameters in the cold room of the cold storage, the data acquisition module is responsible for summarizing and managing the data collected by the sensor module, the data processing module is responsible for analyzing and processing the collected environmental data to support subsequent temperature prediction, and the prediction module is responsible for real-time prediction of future temperature changes in the cold room based on current temperature changes and historical data.

[0009] Preferably, the sensor module includes a temperature sensor and a fan start / stop status sensor. The sensor module includes a temperature sensor installed at multiple key positions in the cold storage to monitor temperature changes in different areas of the cold storage; the fan start / stop status sensor is used to monitor the start / stop status of the fan and record the operating time and status changes of the fan.

[0010] Preferably, the data acquisition module includes a data collector, a data aggregation module, a data synchronization module and a data storage device. The data collector and the data aggregation module are used to receive data collected from various sensors in real time and aggregate them to the data acquisition module via a wired manner; the data synchronization module and the data storage device are used to time synchronize the data from different sensors and store the data.

[0011] Preferably, the main tasks of the data processing module include data cleaning, data preprocessing, feature extraction, and temperature rise and temperature drop segment data segmentation and mathematical modeling based on the start and stop status of the fan.

[0012] A method for cold storage temperature prediction based on mathematical modeling and fitting, comprising the following steps:

[0013] S1. Deploy sensors and data collection. Deploy temperature sensors at multiple key locations in the cold storage, set fan start / stop status sensors on the fans, collect data from each sensor through data collectors, and then aggregate the data from different sensors to the data storage device through wired mode, and synchronize the data from different sensors;

[0014] S2, data cleaning and preprocessing, performing data cleaning and preprocessing on the data collected in step S1 by noise filtering, outlier detection and processing, and data repair;

[0015] S3, data segmentation, using sensors to record the start and stop status of the fan, obtain the accurate time point of the state change, and divide the temperature change data into the temperature rise section and the temperature drop section;

[0016] S4, establish a model and fit, perform mathematical modeling and fitting on the data of each heating section and cooling section, obtain a historical data model and a real-time data model, and obtain the best fitting parameters of each historical temperature change section and the temperature change trend and fitting parameters of the current section respectively;

[0017] S5. Prediction evaluation: By combining the historical data model and the real-time data model and fitting parameters, the historical data model is predicted in the initialization stage, and the real-time data model is dynamically adjusted after the data volume increases.

[0018] Preferably, in step S2, noise filtering adopts low-pass filtering or sliding average filtering, outlier detection and processing adopts statistical methods and range detection methods based on historical data, and data repair adopts linear interpolation, spline interpolation or regression methods based on historical data.

[0019] Preferably, in step S3, the data segmentation process includes the following steps:

[0020] Condition monitoring: Use sensors to record the start and stop status of the fan and obtain the exact time point of the status change;

[0021] Temperature rise period identification: when the fan is turned off or the cold storage door is opened, the time period in which the temperature rises is recorded as the temperature rise period, and the temperature data in this time period is extracted for modeling;

[0022] Identification of cooling section: When the fan is started or the cold storage door is closed, the time period in which the temperature drops is recorded as the cooling section, and the temperature data within this time period is extracted for modeling.

[0023] Preferably, in step S4, the specific steps of historical data modeling are as follows:

[0024] Using historical data, a heating section model and a cooling section model are established for each heating section and cooling section, and the particle swarm optimization algorithm is used to adjust the model parameters to obtain the best fitting parameters for each temperature change section;

[0025] The temperature rise process of the cold room in the mathematical model of the heating section can be divided into two stages: first, the rapid temperature rise stage, followed by the slow temperature rise stage. The mathematical formula is as follows:

[0026]

[0027] Among them, ζ is the acceleration correlation coefficient of temperature change caused by cold diffusion, λ is the temperature attenuation rate coefficient caused by cold diffusion, k is the heat exchange rate between the cold room and the outside world, T0 is the starting temperature, T e It is the upper limit of the cold room temperature control;

[0028] Mathematical model of cooling section is established. After starting a certain number of fans, the temperature of the cold room will drop rapidly, and the speed of temperature drop is positively correlated with the number of fans. As the temperature gradually decreases, the drop rate will slow down, showing obvious attenuation characteristics, and the temperature will eventually stabilize. The mathematical formula is as follows:

[0029]

[0030] T(t)=(T0-T p ) -Nμt +T p

[0031] Among them, T p is the evaporating temperature, T0 is the predicted starting temperature, N is the number of fans turned on, and μ is the cooling capacity coefficient of a single fan.

[0032] Preferably, in step S5, the prediction in the initialization phase includes the following steps:

[0033] First, the prediction module selects historical temperature change segments with similar conditions to the current temperature change segment from historical data, including the number of fans turned on and the initial temperature;

[0034] After matching similar historical data, the system uses the mathematical models and fitting parameters of these data segments as initial prediction tools. At this time, the prediction mainly relies on these historical models to estimate future temperature change trends.

[0035] Using these initial models and parameters, the system makes a preliminary prediction of the temperature changes in the cold room in the short term in the future.

[0036] Preferably, in step S5, the specific steps of dynamic adjustment after the amount of data increases are as follows:

[0037] The system fits the data of the current temperature change section in real time and generates real-time fitting parameters. The above parameters directly reflect the change trend of the current temperature section;

[0038] The real-time fitting parameters of the current temperature segment are weighted averaged with the fitting parameters of the same type of temperature segment in the historical data. In the initial stage, the historical parameters account for the main weight. As time goes by and the amount of data increases, the weight of the current fitting parameters gradually increases.

[0039] During the weighting process, the system dynamically adjusts the weights based on the credibility and quantity of the current data.

[0040] Therefore, the present invention adopts a cold storage temperature prediction system and method based on mathematical modeling and fitting of the above structure, which has the following beneficial effects:

[0041] (1) The present invention adopts fitting and particle swarm optimization algorithms to achieve accurate real-time prediction of cold room temperature changes. It effectively solves the problem that traditional physical mechanism models cannot be accurately modeled due to the large number of complex influencing factors and complex interactions, overcomes the lag of temperature changes, makes the prediction results more timely and accurate, and thus greatly improves the reliability of cold storage temperature control.

[0042] (2) The present invention can accurately predict the temperature change when different numbers of fans are actively turned on or off during any temperature change process, at any time and at any starting temperature. This solves the limitation that the existing AI time series prediction method can only predict periodic patterns. The present invention can dynamically adapt to various operating conditions in the cold storage, provide accurate temperature change predictions, and provide a reliable basis for optimizing cold storage operation strategies and energy-saving control.

[0043] (3) The present invention is very accurate in predicting temperature changes, especially the temperature rise stage. By accurately predicting the temperature rise process, the present invention can effectively control temperature changes, ensure the stability and consistency of the temperature in the cold storage, and avoid the impact of temperature fluctuations on the quality of stored items.

[0044] (4) The present invention is simple in design, easy to integrate and maintain. By accurately predicting the temperature change in the cold room, the present invention can optimize the fan operation strategy, thereby effectively improving energy efficiency. It has strong adaptability and can adapt to different cold storage environments and changes in various influencing factors.

[0045] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A schematic diagram of a cold storage temperature prediction system and method based on mathematical modeling and fitting according to the present invention;

[0047] Figure 2 A schematic diagram of temperature rise of a cold storage temperature prediction system and method based on mathematical modeling and fitting according to the present invention;

[0048] Figure 3 A schematic diagram of temperature drop of a cold storage temperature prediction system and method based on mathematical modeling and fitting according to the present invention;

[0049] Figure 4 A schematic diagram of a cooling process described by different numbers of fans in a cold storage temperature prediction system and method based on mathematical modeling and fitting according to the present invention;

[0050] Figure 5 The present invention is a schematic diagram of dynamic adjustment of a cold storage temperature prediction system and method based on mathematical modeling and fitting after the amount of data increases. DETAILED DESCRIPTION

[0051] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0052] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0053] Example

[0054] like Figure 1 As shown, the present invention provides a cold storage temperature prediction system based on mathematical modeling and fitting, including a sensor module, a data acquisition module, a data processing module and a prediction module. The sensor module, the data acquisition module, the data processing module and the prediction module are connected in sequence. The sensor module is responsible for real-time monitoring of environmental parameters in the cold room of the cold storage. The data acquisition module is responsible for summarizing and managing the data collected by the sensor module. The data processing module is responsible for analyzing and processing the collected environmental data to support subsequent temperature prediction. The prediction module is responsible for real-time prediction of future temperature changes in the cold room based on current temperature changes and historical data.

[0055] The sensor module includes a temperature sensor and a fan start / stop status sensor. The sensor module includes a temperature sensor installed at multiple key locations in the cold storage to monitor temperature changes in different areas of the cold storage; the fan start / stop status sensor is used to monitor the start / stop status of the fan and record the operating time and status changes of the fan.

[0056] The data acquisition module includes a data collector, a data aggregation module, a data synchronization module and a data storage device. The data collector and the data aggregation module are used to receive the data collected from each sensor in real time and aggregate them to the data acquisition module through a wired manner; the data synchronization module and the data storage device are used to synchronize the time of data from different sensors and store the data.

[0057] The main tasks of the data processing module include data cleaning, data preprocessing, feature extraction, and data segmentation and mathematical modeling of the heating and cooling sections based on the start and stop status of the fan.

[0058] A method for cold storage temperature prediction based on mathematical modeling and fitting, comprising the following steps:

[0059] S1. Deploy sensors and data collection. Deploy temperature sensors at multiple key locations in the cold storage, set fan start / stop status sensors on the fans, collect data from each sensor through data collectors, and then aggregate the data from different sensors to the data storage device through wired mode, and synchronize the data from different sensors;

[0060] S2, data cleaning and preprocessing, performing data cleaning and preprocessing on the data collected in step S1 by noise filtering, outlier detection and processing, and data repair;

[0061] S3, data segmentation, using sensors to record the start and stop status of the fan, obtain the accurate time point of the state change, and divide the temperature change data into the temperature rise section and the temperature drop section;

[0062] S4, establish a model and fit, perform mathematical modeling and fitting on the data of each heating section and cooling section, obtain a historical data model and a real-time data model, and obtain the best fitting parameters of each historical temperature change section and the temperature change trend and fitting parameters of the current section respectively;

[0063] S5. Prediction evaluation: By combining the historical data model and the real-time data model and fitting parameters, the historical data model is predicted in the initialization stage, and the real-time data model is dynamically adjusted after the data volume increases.

[0064] In step S2, noise filtering adopts low-pass filtering or sliding average filtering, outlier detection and processing adopts statistical method and range detection method based on historical data, and data repair adopts linear interpolation method, spline interpolation method or regression method based on historical data.

[0065] In step S3, the data segmentation process includes the following steps:

[0066] Condition monitoring: Use sensors to record the start and stop status of the fan and obtain the exact time point of the status change;

[0067] Temperature rise period identification: when the fan is turned off or the cold storage door is opened, the time period in which the temperature rises is recorded as the temperature rise period, and the temperature data in this time period is extracted for modeling;

[0068] Identification of cooling section: When the fan is started or the cold storage door is closed, the time period in which the temperature drops is recorded as the cooling section, and the temperature data within this time period is extracted for modeling.

[0069] In step S4, the specific steps of historical data modeling are as follows:

[0070] Using historical data, a heating section model and a cooling section model are established for each heating section and cooling section, and the particle swarm optimization algorithm is used to adjust the model parameters to obtain the best fitting parameters for each temperature change section;

[0071] like Figure 2 As shown in the figure, the temperature rise process of the cold room in the mathematical model of the heating section can be divided into two stages: first, the rapid temperature rise stage, followed by the slow temperature rise stage. The mathematical formula is as follows:

[0072]

[0073] Among them, ζ is the acceleration correlation coefficient of temperature change caused by cold diffusion, λ is the temperature attenuation rate coefficient caused by cold diffusion, k is the heat exchange rate between the cold room and the outside world, T0 is the starting temperature, T e It is the upper limit of the cold room temperature control;

[0074] like Figure 3 As shown in the figure, the mathematical model of the cooling section is established. After a certain number of fans are started, the temperature of the cold room will drop rapidly, and the speed of temperature drop is positively correlated with the number of fans. As the temperature gradually decreases, the drop rate will slow down, showing obvious attenuation characteristics, and the temperature will eventually stabilize. The mathematical formula is as follows:

[0075]

[0076] T(t)=(T0-T p ) -Nμt +T p

[0077] Among them, T p is the evaporation temperature, T0 is the predicted starting temperature, N is the number of fans turned on, and μ is the cooling capacity coefficient of a single fan. Figure 4 The cooling process described by different numbers of fans is reflected. By fitting and analyzing the temperature change process of the same cold room in different time periods, the obtained model parameters show highly consistent distribution characteristics. This consistency shows that these parameters have clear physical meanings and can accurately reflect the actual physical process.

[0078] In step S5, the prediction in the initialization phase includes the following steps:

[0079] First, the prediction module selects historical temperature change segments with similar conditions to the current temperature change segment from historical data, including the number of fans turned on and the initial temperature;

[0080] After matching similar historical data, the system uses the mathematical models and fitting parameters of these data segments as initial prediction tools. At this time, the prediction mainly relies on these historical models to estimate future temperature change trends.

[0081] Using these initial models and parameters, the system makes a preliminary prediction of the temperature changes in the cold room in the short term in the future.

[0082] The current temperature change process is an unfinished dynamic process. In its initial stage, due to the small amount of data, it is difficult to accurately fit the model, so the model parameters of the same type of temperature change process in the historical data are used for prediction. With the continuous development of the temperature change process and the increase in the amount of data, the fitting modeling results of the current process are gradually taken into consideration to improve the prediction accuracy.

[0083] In the initial stage of temperature change, the empirical model parameters of historical data are fully referenced. As the current temperature change process develops and data accumulates, the weight of the current process fitting model parameters is gradually increased, making the prediction results more and more accurate over time. The specific steps of dynamic adjustment after the amount of data increases are as follows:

[0084] The system fits the data of the current temperature change section in real time and generates real-time fitting parameters. The above parameters directly reflect the change trend of the current temperature section;

[0085] The real-time fitting parameters of the current temperature segment are weighted averaged with the fitting parameters of the same type of temperature segment in the historical data. In the initial stage, the historical parameters account for the main weight. As time goes by and the amount of data increases, the weight of the current fitting parameters gradually increases. This weighted average method ensures that the prediction can not only take advantage of the historical data, but also reflect the characteristics of the current temperature changes in real time.

[0086] During the weighting process, the system dynamically adjusts the weights according to the credibility and quantity of the current data. As the real-time data increases, the weights of the current fitting parameters gradually increase, so that the prediction results more accurately reflect the changing trends of the current segment, such as Figure 5 As shown. P1 (historical data model fitting parameters): represents the parameters of the temperature change model similar to the current process extracted from historical data. These parameters are physically consistent and can better describe the temperature changes under similar conditions. P2 (current temperature change process model fitting parameters): represents the model parameters obtained by fitting the real-time data of the current temperature change process. These parameters can reflect the temperature change law in the current environment in real time. λ (normalized weighting coefficient): is a coefficient used to dynamically adjust the weights of historical parameters and current parameters. Its value is adjusted with the change of time and data volume to ensure that the prediction results not only refer to historical experience but also reflect the current actual changes.

[0087] The entire intelligent prediction system integrates various modules to achieve precise control and management of the cold room temperature in the cold storage:

[0088] System integration: By integrating the data and functions of each module, the overall operation of the system is realized. The modular design is adopted to facilitate the expansion and maintenance of the system.

[0089] Temperature control strategy: According to the prediction results, the fan start and stop status is intelligently adjusted to achieve accurate control of the cold storage temperature. According to different temperature requirements, control strategies can be flexibly formulated to optimize energy consumption.

[0090] Alarm and fault diagnosis: When the temperature prediction results are abnormal, the system can promptly send out an alarm signal and provide fault diagnosis function to help maintenance personnel quickly locate and solve the problem.

[0091] Data storage and analysis: The system can store historical data for a long time and conduct in-depth analysis to provide data support for optimizing cold storage management and formulating energy-saving strategies.

[0092] Therefore, the present invention adopts the above-mentioned cold storage temperature prediction system and method based on mathematical modeling and fitting, and can predict the change of cold room temperature in real time by systematically analyzing and modeling various influencing factors, which is of great significance to the energy-saving automatic control of cold storage. It can not only improve the accuracy of temperature control, but also provide flexible temperature prediction and control strategies under different operating conditions, thereby achieving more efficient cold storage management.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A cold storage temperature prediction system based on mathematical modeling and fitting, It is characterized in that It includes a sensor module, a data acquisition module, a data processing module and a prediction module. The sensor module, the data acquisition module, the data processing module and the prediction module are connected in sequence. The sensor module is responsible for real-time monitoring of environmental parameters in the cold room of the cold storage. The data acquisition module is responsible for summarizing and managing the data collected by the sensor module. The data processing module is responsible for analyzing and processing the collected environmental data to support subsequent temperature prediction. The prediction module is responsible for real-time prediction of future temperature changes in the cold room based on current temperature changes and historical data.

2. According to claim 1, a cold storage temperature prediction system based on mathematical modeling and fitting, Features: The sensor module includes a temperature sensor and a fan start / stop status sensor. The sensor module includes a temperature sensor installed at multiple key positions in the cold storage to monitor temperature changes in different areas of the cold storage; the fan start / stop status sensor is used to monitor the start / stop status of the fan and record the operating time and status changes of the fan.

3. According to claim 1, a cold storage temperature prediction system based on mathematical modeling and fitting, Features: The data acquisition module includes a data collector, a data aggregation module, a data synchronization module and a data storage device. The data collector and the data aggregation module are used to receive data collected from various sensors in real time and aggregate them to the data acquisition module in a wired manner; The data synchronization module and the data storage are used to perform time synchronization on data from different sensors and store the data.

4. According to claim 1, a cold storage temperature prediction system based on mathematical modeling and fitting, Features: The main tasks of the data processing module include data cleaning, data preprocessing, feature extraction, and temperature rise and temperature drop data segmentation and mathematical modeling based on the start and stop status of the fan.

5. A method for cold storage temperature prediction based on mathematical modeling and fitting according to any one of claims 1 to 4, Features: The following steps are involved: S1. Deploy sensors and data collection. Deploy temperature sensors at multiple key locations in the cold storage, set fan start / stop status sensors on the fans, collect data from each sensor through data collectors, and then aggregate the data from different sensors to the data storage device through wired mode, and synchronize the data from different sensors; S2, data cleaning and preprocessing, performing data cleaning and preprocessing on the data collected in step S1 by noise filtering, outlier detection and processing, and data repair; S3, data segmentation, using sensors to record the start and stop status of the fan, obtain the accurate time point of the state change, and divide the temperature change data into the temperature rise section and the temperature drop section; S4, establish a model and fit, perform mathematical modeling and fitting on the data of each heating section and cooling section, obtain a historical data model and a real-time data model, and obtain the best fitting parameters of each historical temperature change section and the temperature change trend and fitting parameters of the current section respectively; S5. Prediction evaluation: By combining the historical data model and the real-time data model and fitting parameters, the historical data model is predicted in the initialization stage, and the real-time data model is dynamically adjusted after the data volume increases.

6. A method for cold storage temperature prediction based on mathematical modeling and fitting according to claim 5, Features: In step S2, noise filtering adopts low-pass filtering or sliding average filtering, outlier detection and processing adopts statistical method and range detection method based on historical data, and data repair adopts linear interpolation method, spline interpolation method or regression method based on historical data.

7. According to claim 5, a method for cold storage temperature prediction based on mathematical modeling and fitting, Features: In step S3, the data segmentation process includes the following steps: Condition monitoring: Use sensors to record the start and stop status of the fan and obtain the exact time point of the status change; Temperature rise period identification: when the fan is turned off or the cold storage door is opened, the time period in which the temperature rises is recorded as the temperature rise period, and the temperature data in this time period is extracted for modeling; Identification of cooling section: When the fan is started or the cold storage door is closed, the time period in which the temperature drops is recorded as the cooling section, and the temperature data within this time period is extracted for modeling.

8. According to claim 5, a method for cold storage temperature prediction based on mathematical modeling and fitting, Features: In step S4, the specific steps of historical data modeling are as follows: Using historical data, a heating section model and a cooling section model are established for each heating section and cooling section, and the particle swarm optimization algorithm is used to adjust the model parameters to obtain the best fitting parameters for each temperature change section; The temperature rise process of the cold room in the mathematical model of the heating section can be divided into two stages: first, the rapid temperature rise stage, followed by the slow temperature rise stage. The mathematical formula is as follows: Among them, ζ is the acceleration correlation coefficient of temperature change caused by cold diffusion, λ is the temperature attenuation rate coefficient caused by cold diffusion, k is the heat exchange rate between the cold room and the outside world, T 0 is the starting temperature, T e It is the upper limit of the cold room temperature control; Mathematical model of cooling section is established. After starting a certain number of fans, the temperature of the cold room will drop rapidly, and the speed of temperature drop is positively correlated with the number of fans. As the temperature gradually decreases, the drop rate will slow down, showing obvious attenuation characteristics, and the temperature will eventually stabilize. The mathematical formula is as follows: Among them, T p is the evaporation temperature, T 0 is the predicted starting temperature, N is the number of fans turned on, and μ is the cooling capacity coefficient of a single fan.

9. A method for cold storage temperature prediction based on mathematical modeling and fitting according to claim 5, Features: In step S5, the prediction in the initialization phase includes the following steps: First, the prediction module selects historical temperature change segments with similar conditions to the current temperature change segment from historical data, including the number of fans turned on and the initial temperature; After matching similar historical data, the system uses the mathematical models and fitting parameters of these data segments as initial prediction tools. At this time, the prediction mainly relies on these historical models to estimate future temperature change trends. Using these initial models and parameters, the system makes a preliminary prediction of the temperature changes in the cold room in the short term in the future.

10. A cold storage temperature prediction system and method based on mathematical modeling and fitting according to claim 9, Features: In step S5, the specific steps of dynamic adjustment after the data volume increases are as follows: The system fits the data of the current temperature change section in real time and generates real-time fitting parameters. The above parameters directly reflect the change trend of the current temperature section; The real-time fitting parameters of the current temperature segment are weighted averaged with the fitting parameters of the same type of temperature segment in the historical data. In the initial stage, the historical parameters account for the main weight. As time goes by and the amount of data increases, the weight of the current fitting parameters gradually increases. During the weighting process, the system dynamically adjusts the weights based on the credibility and quantity of the current data.

Images