Intelligent commercial refrigerator in energy-saving mode based on AI technology

Abstract

The invention discloses an intelligent commercial refrigerator in an energy-saving mode based on an AI technology. The intelligent commercial refrigerator comprises an AI intelligent control module, a multi-dimensional sensor network module, a dynamic refrigeration adjusting module, an intelligent defrosting control module, a monitoring management module, a diagnosis correction module and an energy-saving module. The problems that a traditional commercial refrigerator is high in energy consumption, inflexible in refrigeration adjustment, not timely in fault diagnosis and incapable of achieving remote monitoring are solved. According to the invention, through specific algorithms of the improved sparrow algorithm module and the over-temperature calculation module in the AI intelligent control module, multi-dimensional data in the operation process of the refrigerator are evaluated and processed, and the accuracy and efficiency of data processing are improved. And through an improved time sequence analysis module and a cooling capacity demand prediction algorithm module in the dynamic refrigeration adjustment module, the refrigeration demand of the refrigerator is evaluated and optimized, and real-time adjustment of the refrigeration intensity and the operation state is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent commercial refrigerators, more particularly to an intelligent commercial refrigerator with an energy-saving mode based on AI technology. BACKGROUND

[0002] With the increasing demand for energy-efficient equipment in the commercial field, the energy consumption of intelligent commercial refrigerators, which are commonly used in commercial places, has attracted much attention. However, the existing intelligent commercial refrigerator technology has slow response when the refrigerator faces environmental changes or user operations, and cannot adjust the refrigeration and defrosting functions in time.

[0003] In the prior art, the collection of partial data such as temperature and humidity in the refrigerator has been well implemented, providing a certain data basis for the operation control of the refrigerator. However, the problems of slow control feedback and uneven energy resource management and scheduling are still prominent, seriously affecting the energy-saving effect and use performance of the refrigerator.

[0004] To solve the above problems, the present application sets an AI intelligent control module to analyze and accurately control the multi-dimensional data of the refrigerator operation in real time, solves the problem of slow control feedback of the refrigerator system, and uses a prediction algorithm and a scheduling algorithm to calculate and analyze the energy supply and demand balance of the refrigerator, and reasonably distributes the energy through an energy-saving module to solve the problem of uneven energy resource management and scheduling. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application discloses an intelligent commercial refrigerator with an energy-saving mode based on AI technology. A multi-dimensional sensor network module is set to sense the data of the refrigerator temperature, humidity, door opening times, article pressure and frost layer in real time. An AI intelligent control module is combined with a central processor module and an improved sparrow algorithm module to analyze and control the multi-dimensional data in real time. An improved time series analysis module and a cold load demand prediction algorithm module in the dynamic refrigeration adjustment module are used to adjust the refrigeration intensity in real time according to the feedback of the AI intelligent control module. A compressor control module, a fan control module and a valve control module in the energy-saving module are used to control the temperature of the refrigerator.

[0006] The present application adopts the following technical solutions:

[0007] An intelligent commercial refrigerator system with an energy-saving mode based on AI technology, characterized in that: the intelligent commercial refrigerator is provided with an AI intelligent control module, a multi-dimensional sensor network module, a dynamic refrigeration adjustment module, an intelligent defrosting control module, a monitoring management module, a diagnosis correction module and an energy-saving module; wherein:

[0008] The AI intelligent control module controls multi-dimensional data in the operation process of the refrigerator in real time, and comprises a central processor module, a data acquisition module connected with the central processor module, an improved sparrow algorithm module, an over-temperature calculation module, and a data output interface.

[0009] The multi-dimensional sensor network module is used for sensing refrigerator temperature, humidity, door opening times, article pressure, and frost layer data, and is provided with temperature sensors, humidity sensors, weight sensors, pressure sensors, and light sensors arranged inside and outside the refrigerator.

[0010] The dynamic refrigeration adjustment module adjusts refrigeration strength and operation state in real time according to feedback data of the AI intelligent control module, and comprises a database unit transmitted by the AI intelligent control module, a data processing module, an improved time series analysis module, a cold quantity demand prediction algorithm module, and an execution mechanism.

[0011] The intelligent defrosting control module performs intelligent defrosting by monitoring the frosting condition of the evaporator in real time through sensors, and comprises a frost layer sensor, a sensor data fusion module, a control execution unit, a heating control module, and a drainage control module.

[0012] The monitoring management module connects the refrigerator intelligent control system through the Internet to view the operation state of the refrigerator in real time, and comprises a communication module, an information collection module connected with the communication module, a data arrangement module, and a remote monitoring system.

[0013] The diagnosis correction module predicts and diagnoses potential faults, and comprises a fault diagnosis model and a self-repair execution unit.

[0014] The energy-saving module performs energy saving and consumption reduction of the refrigerator according to data transmitted by the AI intelligent control module.

[0015] The output end of the multi-dimensional sensor network module, the output end of the dynamic refrigeration adjustment module, the output end of the intelligent defrosting control module, the output end of the monitoring management module, and the output end of the energy-saving module are connected with the input end of the AI intelligent control module; the output end of the AI intelligent control module is connected with the input end of the dynamic refrigeration adjustment module, the input end of the intelligent defrosting control module, the input end of the monitoring management module, the input end of the multi-dimensional sensor network module, and the input end of the energy-saving module; the output end of the monitoring management module and the output end of the intelligent defrosting control module are connected with the input end of the diagnosis correction module; the output end of the energy-saving module is connected with the input end of the dynamic refrigeration adjustment module; and the output end of the diagnosis module is connected with the input end of the AI intelligent control module.

[0016] As a further technical solution of the present invention, the working method of the AI ​​intelligent control module is as follows:

[0017] Step 1: Obtain the freezer's operating status data.

[0018] The data acquisition module dynamically acquires data on the temperature, humidity, pressure, current, voltage, door open / close status, and light intensity of the freezer. After collecting and processing the data, the data acquisition module transmits it to the data processing and analysis unit.

[0019] Step 2: Calculate the optimal temperature using the improved Sparrow Algorithm module.

[0020] The improved sparrow algorithm module uses fan speed and refrigeration valve opening as optimization variables, with the goal of minimizing temperature fluctuations, to optimize the parameters of fan speed and refrigeration valve opening and control the temperature accurately.

[0021] Step 3: Over-temperature calculation module

[0022] The temperature calculation module determines whether the equipment temperature exceeds the safe or normal operating range and performs statistical analysis on over-temperature situations. Its working method is as follows:

[0023] Based on the equipment's operating requirements and safety standards, a temperature threshold is preset, setting the over-temperature threshold of the cold storage area to 5°C. The processing unit receives temperature data collected by the temperature sensor in real time and compares it with the set threshold. If the current temperature exceeds the threshold, the equipment is determined to be in an over-temperature state; otherwise, it is in a normal temperature state.

[0024] Step 4: Through the data output interface

[0025] The integrated data is sent to the compressor in energy-saving mode via the data output module.

[0026] As a further technical solution of the present invention, the temperature sensor monitors the temperature of different areas inside the freezer in real time to ensure that the temperature is within a suitable range; the humidity sensor detects the humidity inside the freezer; the light sensor is installed inside the freezer, and when the door is opened, the light sensor detects the change in light to determine the opening status of the door; the pressure sensor is installed on the pipes of the refrigeration system to monitor the pressure change of the refrigerant; and the weight sensor is installed on the shelves of the freezer to monitor the weight change of the goods inside the freezer in real time.

[0027] As a further technical solution of the present invention, the working method of the dynamic cooling regulation module is as follows:

[0028] Step 1: Store the freezer data information through database units;

[0029] The sensors continuously collect data on the internal temperature, humidity, pressure, weight of items, and external ambient temperature of the freezer, and transmit them to the database unit. The database unit cleans, organizes, and analyzes the collected data, identifies outliers and noise, and processes them accordingly.

[0030] Step 2: Analyze and calculate the input data using the improved time series analysis module.

[0031] Using an improved time series analysis combined with multiple regression, the time series analysis portion uses an improved autoregressive moving average model to predict the trend term Q of cooling demand. 趋势 :

[0032]

[0033] In formula (1), Q 趋势 Q represents the trend of cooling demand. 历史 This represents historical cooling demand data, where p represents the autoregressive order, q represents the moving average order, and r represents the cooling demand growth order. θ represents the coefficient corresponding to the autoregressive order. j δ represents the coefficient corresponding to the order of the moving average. z Indicates the increase in cooling capacity, ∈ t-j This indicates a past error term;

[0034] An improved formula for cooling demand is established through multiple regression analysis:

[0035]

[0036] In formula (2), Q 需求 (t) represents the cooling demand, β0 represents a constant term, and β1-β3 are the regression coefficients of each factor. 环境 (t) represents the ambient temperature, H 湿度 (t) represents humidity, N 开门次数 (t) represents the number of times the door is opened, β i ρ represents the sum of regression coefficients. i Indicates the total quantity of goods;

[0037] The improved final cooling demand forecast is calculated using the trend term and a weighted sum of multiple regression equations:

[0038] Q 预测 (t)=αQ 趋势 (t)+βQ 需求 (t)+(1-αβ)Q 趋势 (t)Q 需求 (t) (3)

[0039] In formula (3), Q 预测(t) represents the final cooling capacity demand prediction value, Q 趋势 (t) represents the cooling capacity demand trend item based on time series analysis, Q 需求 (t) represents the cooling capacity demand prediction value based on multiple regression, and a represents the weight coefficient of the trend item, wherein 0≤a≤1, and β represents the weight and coefficient of the regression phase;

[0040] Step three, distributing the control instructions through the dynamic refrigeration adjustment module

[0041] The database unit converts the prepared refrigeration adjustment strategy into specific control instructions and sends them to the control execution unit; after receiving the instructions, the control execution unit adjusts the working state of the compressor through the compressor controller, adjusts the fan speed through the fan controller, and controls the electromagnetic valve to act through the electromagnetic valve controller, thereby jointly realizing the dynamic adjustment of the refrigeration system of the refrigerator; the dynamic refrigeration adjustment algorithm is to adjust the compressor speed and the fan power, and the improved compressor speed adjustment formula is:

[0042]

[0043] In formula (4), N(t) represents the target speed of the compressor at time t, N 基础 represents the basic speed of the compressor, e(t) represents the error signal, the deviation between the current temperature and the target temperature, K p represents the proportional gain coefficient, K i represents the integral gain coefficient, and K d represents the differential gain coefficient.

[0044] The improved fan power adjustment output function is:

[0045]

[0046] In formula (5), P(t) represents the target speed of the compressor at time t, P 基础 represents the basic speed of the compressor, e(t) represents the error signal, the deviation between the current temperature and the target temperature, K p represents the proportional gain coefficient, K i represents the integral gain coefficient, and K d represents the differential gain coefficient.

[0047] As a further technical solution of the present application, the working method of the intelligent defrosting control module is:

[0048] The frost layer sensor collects data in the refrigerator and transmits the data to the sensing data fusion module; the sensing data fusion module determines defrosting and sends a defrosting instruction to the control execution unit; after the control execution unit receives the defrosting instruction, the heating control module starts the heating device of the evaporator, heats the evaporator according to the set heating power, and melts the frost layer; the compressor and the fan control module adjust the operating state of the compressor and the fan according to the defrosting requirement; the drain control module opens the drain valve or starts the drain pump after the frost layer is melted to drain the melted water out of the refrigerator.

[0049] As a further technical solution of the application, the working method of the monitoring management module is:

[0050] The monitoring management module monitors the running state of the refrigerator in real time; the communication module receives various data from the intelligent control system of the refrigerator, the information collection module stores the data, the data processing module cleans, classifies and analyzes the data, and the remote monitoring system provides an intuitive interface for the user to view the running state of the refrigerator and can perform corresponding control operations according to requirements.

[0051] As a further technical solution of the application, the working method of the diagnosis correction module is:

[0052] The collected data is preprocessed, and the processed data is input into the fault diagnosis model; the fault diagnosis model compares and analyzes the fault mode in the knowledge base to determine whether the refrigerator has a fault, the type and severity of the fault; according to the fault diagnosis result, the control module determines whether the fault can be solved by self-repairing measures.

[0053] As a further technical solution of the application, the working method of the energy-saving module is:

[0054] The temperature, humidity, pressure, current, voltage and door switch state of the refrigerator are collected in real time, and the data processing module pre-processes the received data and inputs it into the energy-saving strategy model; the execution and control part controls the compressor, fan and valve components of the refrigerator according to the energy-saving strategy through the compressor control module, fan control module and valve control module, and adjusts the operating parameters of the refrigerator.

[0055] Compared with the prior art, the application has the beneficial positive effects that:

[0056] The application sets up a multi-dimensional sensor network module to sense and collect multi-dimensional data of temperature, humidity, door opening times, article pressure and frost layer inside and outside the refrigerator in real time; through the collaborative work of AI intelligent control module, dynamic refrigeration adjustment module, intelligent defrosting control module, monitoring management module, diagnosis correction module and energy saving module, the precise control of the refrigerator running state is realized; to solve the problem of energy consumption of the refrigerator, so as to ensure the safe and stable operation of the refrigerator in the energy saving mode, effectively reduce the energy consumption, improve the refrigeration and preservation effect, and prolong the service life of the equipment. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained according to these drawings without creative labor for those skilled in the art.

[0058] Figure 1 The overall architecture schematic diagram of the intelligent commercial refrigerator system of the energy saving mode based on AI technology;

[0059] Figure 2 The working method flow chart of the AI intelligent control module of the intelligent commercial refrigerator system of the energy saving mode based on AI technology;

[0060] Figure 3 The working method flow chart of the dynamic refrigeration adjustment module of the intelligent commercial refrigerator system of the energy saving mode based on AI technology;

[0061] Figure 4 The working method flow chart of the intelligent defrosting control module of the intelligent commercial refrigerator system of the energy saving mode based on AI technology;

[0062] Figure 5 The working method flow chart of the monitoring management module of the intelligent commercial refrigerator system of the energy saving mode based on AI technology. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. It should be understood that the description is only exemplary, and is not intended to limit the scope of the present application. In addition, in the description, the description of the known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present application.

[0064] As Figure 1As shown, an AI technology-based energy-saving mode intelligent commercial refrigerator system is characterized in that: the intelligent commercial refrigerator table is provided with: an AI intelligent control module, a multi-dimensional sensor network module, a dynamic refrigeration adjustment module, an intelligent defrosting control module, a monitoring management module, a diagnosis correction module and an energy-saving module; wherein:

[0065] The AI intelligent control module controls and analyzes multi-dimensional data in the refrigerator running process in real time, and includes a central processor module, a data acquisition module connected with the central processor module, an improved sparrow algorithm module, an over-temperature calculation module and a data output interface.

[0066] The multi-dimensional sensor network module is used for sensing refrigerator temperature, humidity, door opening times, article pressure and frost data, and is provided with temperature sensors, humidity sensors, weight sensors, pressure sensors and light sensors inside and outside the refrigerator.

[0067] The dynamic refrigeration adjustment module adjusts the refrigeration intensity and operating state in real time according to the feedback data of the AI intelligent control module, and includes a database unit transmitted by the AI intelligent control module, a data processing module, an improved time series analysis module, a cold demand prediction algorithm module and an execution mechanism.

[0068] The intelligent defrosting control module intelligently defrosts by real-time monitoring of the evaporator frosting condition through sensors, and includes a frost sensor, a sensor data fusion module, a control execution unit, a heating control module and a drainage control module.

[0069] The monitoring management module can view the refrigerator running state in real time through the internet connection of the refrigerator intelligent control system, and includes a communication module, a database connected with the communication module, a data arrangement module and a remote monitoring system.

[0070] The diagnosis correction module predicts and diagnoses potential faults, and includes a fault diagnosis model and a self-repair execution unit.

[0071] The energy-saving module performs refrigerator energy-saving and consumption reduction according to the data transmitted by the AI intelligent control module.

[0072] The output end of the multi-dimensional sensor network module, the output end of the dynamic refrigeration adjustment module, the output end of the intelligent defrosting control module, the output end of the monitoring management module, and the output end of the energy-saving module are connected to the input end of the AI intelligent control module. The output end of the AI intelligent control module is connected to the input end of the dynamic refrigeration adjustment module, the input end of the intelligent defrosting control module, the input end of the monitoring management module, the input end of the multi-dimensional sensor network module, and the input end of the energy-saving module. The output end of the monitoring management module, the output end of the intelligent defrosting control module, and the output end of the dynamic refrigeration adjustment module are connected to the input end of the diagnostic correction module. The output end of the energy-saving module is connected to the input end of the dynamic refrigeration adjustment module. The output end of the diagnostic module is connected to the input end of the AI intelligent control module.

[0073] Further, the working method of the AI intelligent control module is:

[0074] Step one, obtain refrigerator working state data information

[0075] The data acquisition module dynamically acquires the data of temperature, humidity, pressure, current, voltage, door switch state, and light intensity of the refrigerator. After the data acquisition module collects and processes the data, it transmits the data to the data processing and analysis unit.

[0076] The data acquisition module plays a crucial role in the refrigerator monitoring system, and its working process is as follows:

[0077] During data acquisition, the data acquisition capability is improved through sensor configuration, such as temperature sensors, humidity sensors, pressure sensors, current sensors, voltage sensors, door state sensors, and light intensity sensors, in order to collect various parameters of the refrigerator. The sensors monitor the environmental parameters inside and outside the refrigerator in real time, such as temperature, humidity, pressure, and electrical parameters such as current, voltage, as well as the opening and closing state of the refrigerator door and the light intensity.

[0078] During data collection, signal conversion is used to convert physical quantities (such as temperature, humidity, etc.) into electrical signals by the sensor. The data acquisition module reads the electrical signals of the sensor through an analog-to-digital converter (ADC) or other conversion mechanisms and converts these signals into digital data. In order to eliminate noise and interference, the data acquisition module filters the collected data to ensure the accuracy of the data. According to the calibration data of the sensor, the collected data is calibrated to eliminate the inherent errors of the sensor. The raw data is converted into the required format, such as Celsius, percentage, etc., to facilitate subsequent processing and analysis. If the data volume is large, data compression is needed to reduce the transmission and storage requirements. The data acquisition module contains local storage capability for temporary storage of collected data. The data acquisition module transmits the processed data to the data processing and analysis unit through wireless communication (such as Wi-Fi, Bluetooth, ZigBee, etc.) or wired communication (such as Ethernet, RS-485, etc.). The data processing and analysis unit receives the data transmitted by the data acquisition module. The collected data is analyzed, including trend analysis, anomaly detection, performance evaluation, etc. According to the analysis results, the data processing and analysis unit generates alarms, optimization suggestions or operation instructions. If an anomaly is detected or optimization is needed, the data processing and analysis unit will send control signals to the refrigerator control system to adjust the operating parameters of the refrigerator, such as adjusting the refrigeration temperature, adjusting the fan speed, etc. The system will send alarm information to relevant personnel through email, SMS or other notification methods.

[0079] This process ensures that the operating status of the refrigerator is always under monitoring and can be timely responded and adjusted to maintain the best working condition.

[0080] Step two, calculate the optimal temperature through the improved sparrow algorithm module

[0081] The improved sparrow algorithm module takes the fan speed and refrigeration valve opening as optimization variables, minimizes the temperature fluctuation as the target, optimizes the fan speed and refrigeration valve opening parameters, and controls the precision of the temperature;

[0082] In this step, the improved Sparrow Algorithm (SIS Algorithm) is an optimization algorithm that simulates the foraging and migration behavior of a Sparrow population to find the optimal solution. In the optimization scenario of fan speed and refrigeration valve opening, the working process of the improved Sparrow Algorithm module is as follows: the algorithm initializes a Sparrow population, and each individual in the population represents a combination of fan speed and refrigeration valve opening parameters. The position of each Sparrow is determined by the values of fan speed and refrigeration valve opening, and the speed represents the rate of parameter change. For each Sparrow in the population, the algorithm uses the objective function (temperature fluctuation minimization) to evaluate its performance. Sparrows forage according to individual performance and group experience, i.e., search for better parameter combinations. Sparrows perform random searches near their current positions to find better solutions. Sparrows also perform global searches based on the position of the best individual in the group to escape local optima. The improved Sparrow Algorithm introduces dynamic adjustment mechanisms, such as adjusting the search range and speed based on the number of iterations or individual performance. To maintain the diversity of the population, the algorithm introduces some strategies, such as random disturbance, elite retention, etc., to avoid premature convergence to local optima. After the algorithm iteration is completed, the position of the best individual in the population corresponds to the optimized fan speed and refrigeration valve opening parameters. Finally, the control output applies these parameters to the refrigerator control system to achieve precise temperature control. Apply the optimized parameters to the actual system and collect feedback data. According to the feedback data, the optimization algorithm needs to be run again to further optimize the parameters.

[0083] Through this process, the improved Sparrow Algorithm module can effectively find the best parameter combination of fan speed and refrigeration valve opening, thereby minimizing temperature fluctuations and improving the efficiency and stability of the refrigeration system.

[0084] Step three, over-temperature calculation module

[0085] The over-temperature calculation module determines whether the device temperature exceeds the safe or normal operating range and performs statistics and analysis on the over-temperature situation. Its working method is as follows:

[0086] According to the working requirements and safety standards of the device, a temperature threshold is set in advance, and the over-temperature threshold of the refrigeration area is set. The processing unit will receive the temperature data collected by the temperature sensor in real time and compare it with the set threshold. If the current temperature exceeds the threshold, it is determined that the device is in an over-temperature state. Otherwise, it is in a normal temperature state.

[0087] Step four, output data through the data output interface

[0088] The integrated data is sent to the compressor in energy-saving mode through the data output module.

[0089] In specific embodiments, the data acquisition module is equipped with high-precision temperature sensors, with an accuracy of ±0.1°C in the refrigeration area and ±0.3°C in the freezing area, and data is collected every 5 seconds; the humidity sensor can accurately measure the humidity in the refrigerator, and data is collected every 10 seconds; the pressure sensor is used to detect the pressure when the door is closed to determine whether the door is tightly closed, and data is collected every time the door is opened or closed. In the running process of the improved sparrow algorithm module, through continuous iteration optimization, it is determined that when the refrigeration area has a large daytime passenger flow and the door is frequently opened and closed, the fan speed is adjusted to 1200 rpm and the refrigeration valve opening is 60%, which can control the temperature fluctuation in the smallest range; while in the small passenger flow period at night, the fan speed is reduced to 800 rpm and the refrigeration valve opening is adjusted to 40%, which can maintain temperature stability and effectively save energy. In terms of the over-temperature calculation module, the over-temperature threshold of the refrigeration area is set to 6°C, and when the temperature sensor detects that the temperature exceeds the threshold for 3 times in a row, it is determined that the over-temperature state, the system immediately starts the alarm mechanism, sends alarm information to the manager's mobile phone APP, and records the over-temperature start time, end time and maximum temperature and other data. The output module sends various types of data processed and analyzed, such as real-time temperature, humidity, optimized fan speed and refrigeration valve opening parameters, and over-temperature statistical information, to the compressor control system in a wireless transmission manner.

[0090] Further, the temperature sensor monitors the temperature of different areas inside the refrigerator in real time to ensure that the temperature is within a suitable range; the humidity sensor detects the humidity in the refrigerator; the light sensor is installed inside the refrigerator, and when the door is opened, the light sensor detects changes in light to determine the opening state of the door; the pressure sensor is installed on the pipeline of the refrigeration system to monitor the pressure change of the refrigerant; the weight sensor is installed on the shelf of the refrigerator to monitor the weight change of the goods in the refrigerator in real time.

[0091] In specific embodiments, high-precision temperature sensors are installed in the refrigeration area to accurately sense temperature changes and ensure that the temperature of the entire refrigeration area is maintained between 2°C and 5°C. Once the temperature exceeds this range, the temperature sensor quickly feeds data back to the AI intelligent control module, which then instructs the dynamic refrigeration adjustment module to adjust the refrigeration intensity. The humidity sensor is installed on the back wall of the refrigerator near the air outlet, which can monitor the humidity in the refrigerator in real time and transmit the data to the AI intelligent control module. When the humidity is lower than 85%, the intelligent control module can instruct the humidifying device in the refrigerator to start to increase the humidity; if the humidity is higher than 95%, the humidity can be reduced by increasing ventilation, etc. The light sensors are evenly distributed in the refrigerator. When the customer opens the door, the light changes instantly, and the light sensor immediately captures this signal and quickly transmits it to the AI intelligent control module. After receiving the signal, the module can control the lighting system in the refrigerator to turn on, making it convenient for customers to select goods. The pressure sensor is installed on the high-pressure pipe and low-pressure pipe of the refrigeration system to monitor the pressure changes of the refrigerant in real time. Under normal operation, the pressure of the high-pressure pipe is 2.0 MPa, and the pressure of the low-pressure pipe is 0.4 MPa. Once the pressure fluctuates abnormally, the pressure sensor feeds the data back to the diagnostic correction module, which quickly analyzes and judges the fault causes, such as whether there is a refrigerant leak and compressor failure, and takes timely measures. The weight sensor is installed at the bottom of each layer of shelves to accurately measure the weight of goods on the shelves.

[0092] Further, the working method of the dynamic refrigeration adjustment module is as follows:

[0093] Step 1: Store the refrigerator data information through the database unit;

[0094] The sensors continuously collect data of the internal temperature, humidity, pressure, and weight of goods in the refrigerator, as well as the external environmental temperature of the refrigerator, and transmit them to the database unit. The database unit cleans, organizes, and analyzes the collected data, identifies and processes abnormal values and noise in the data;

[0095] Step 2: Analyze and calculate the data input through the improved time series analysis module to calculate the cooling demand

[0096] Using the improved time series analysis combined with the method of multiple regression, the time series analysis part predicts the trend item Q of the cooling demand through the improved autoregressive moving average model 趋势 :

[0097]

[0098] In formula (1), Q 趋势 represents the trend item of the cooling demand, Q 趋势 represents the historical cooling demand data, p represents the autoregressive order, q represents the moving average order, and r represents the cooling growth order, denotes the coefficient corresponding to the autoregressive order, θ j denotes the coefficient corresponding to the moving average order, δ z denotes the amount of cold energy growth, ∈ t-j denotes the past error term;

[0099] Equation (1) describes an improved autoregressive moving average model (ARIMA) for predicting the trend item of cold energy demand. The following is the specific working process of this model in the time series analysis part:

[0100] First, collect historical cooling demand data, which is usually in the form of time series, such as daily, weekly, or monthly cooling demand. Clean the collected data, including handling missing values, outliers, and seasonal adjustment. Use unit root tests (such as ADF test) to check the stationarity of the time series data. If the data is non-stationary, it needs to be differenced. Determine the autoregressive order (p) and moving average order (q) by using autocorrelation plots (ACF) and partial autocorrelation plots (PACF). These plots show the autocorrelation and partial autocorrelation patterns of the data series. Estimate the model parameters using maximum likelihood estimation (MLE) or other methods. For ARIMA models, estimate the autoregressive coefficients (φ), moving average coefficients (θ), and cooling growth coefficients (β). Determine the cooling growth order (d) if the data is differenced, which determines how many times the data needs to be differenced to achieve stationarity. Substitute the estimated parameters into equation (1) to fit the modified autoregressive moving average model. Use residual analysis to check the fitting effect of the model, including the serial correlation of residuals, constant variance, and normality. The trend term (T_t) is the trend term of the model's predicted cooling demand, reflecting the trend of cooling demand over time. The historical cooling demand data (X_t) is the current observation in the time series. The autoregressive order (p) is the order of the autoregressive term, representing the relationship between the current observation and the past p observations. The moving average order (q) is the order of the moving average term, representing the relationship between the current observation and the moving average of the past q observations. The cooling growth order (d) is the order of the difference, representing how many times the data needs to be differenced to achieve stationarity. The autoregressive coefficient (φ) represents the weight of the autoregressive term, indicating the influence of past observations on the current observation. The moving average coefficient (θ) represents the weight of the moving average term, indicating the influence of the moving average of past observations on the current observation. The cooling growth amount (β) represents the growth trend of cooling demand. The error term (ε_t) is the model residual, representing random fluctuations that the model cannot explain. Use the fitted model to predict future cooling demand trends. By analyzing the prediction results and adjusting the model parameters or taking appropriate measures according to the actual situation. Through this process, the modified autoregressive moving average model can predict the trend of cooling demand, providing data support for energy management and control of refrigerators or other refrigeration systems.

[0101] The multiple regression part establishes a modified cooling demand formula:

[0102]

[0103] In equation (2), Q 需求 (t) represents the cooling demand, β0 represents the constant term, and β1-β3 are the regression coefficients of each factor, T 环境(t) represents the ambient temperature, H 湿度 (t) represents the humidity, N 开门次数 (t) represents the number of door openings, β i represents the sum of regression coefficients, ρ i represents the total number of goods;

[0104] In the specific work, formula (2) describes a multiple regression model for establishing the relationship between the cooling demand and multiple influencing factors. The following is the working process of establishing an improved cooling demand formula. First, collect data related to cooling demand, including historical cooling demand and factors that affect cooling demand, such as ambient temperature, humidity, door opening frequency, and total number of goods. Clean the collected data, handle missing values and outliers, and perform data conversion (such as normalization or standardization). Then determine the independent variables, according to domain knowledge and data analysis, select the factors that affect the cooling demand as independent variables (X1, X2, X3,...). When controlling variables, some control variables need to be considered to eliminate their influence on cooling demand.

[0105] Set the general form of the multiple regression model, that is, the form of formula (2), where Y is the dependent variable (cooling demand), β0 is the constant term, βi is the regression coefficient of each factor, and Xi is the independent variable (such as ambient temperature, humidity, door opening frequency, etc.). Use statistical software or programming tools (such as R, Python, etc.) to perform multiple regression analysis and estimate model parameters (β0, β1, β2,...). Perform significance tests on the estimated regression coefficients, such as t-test or F-test, to determine which factors have a significant impact on cooling demand. When analyzing residuals, first check the distribution of model residuals to ensure they are random, without autocorrelation and heteroscedasticity.

[0106] When checking for multicollinearity, check whether there is a high correlation between independent variables, which leads to an unstable model. According to formula (2), set up a multiple regression model with cooling demand as the dependent variable and ambient temperature, humidity, door opening frequency, etc. as independent variables. Use the least squares method or other optimization methods to estimate model parameters. Explain the meaning of each regression coefficient, for example, the coefficient of ambient temperature represents the change in cooling demand when ambient temperature changes by one unit. Apply the model to new data sets to evaluate the predictive ability of the model. Use cross-validation or leave-out methods to verify the predictive performance of the model. Use the established model to predict future cooling demand. Use the prediction results to optimize inventory management, energy consumption, and refrigeration system design.

[0107] Through the above process, an improved cooling demand formula can be established, which can help predict and analyze key factors affecting cooling demand, providing a scientific basis for the optimization and management of refrigeration systems.

[0108] The improved final cooling demand prediction value is a combination of the trend term and the multiple regression weighted sum formula:

[0109] Q 预测 (t) = αQ 趋势 (t) + βQ 需求 (t) + (1 - αβ)Q 趋势 (t)Q 需求 (t) (3)

[0110] In formula (3), Q 预测 (t) represents the time final cooling demand prediction value, Q 趋势 (t) represents the cooling demand trend term based on time series analysis, Q 需求 (t) represents the cooling demand prediction value based on multiple regression, α represents the weight coefficient of the trend term, where 0 ≤ α ≤ 1, and β represents the weight and coefficient of the regression phase.

[0111] Formula (3) combines time series analysis and multiple regression methods to predict the final cooling demand value. This method combines the results of two different prediction models by weighted average, in order to obtain more accurate prediction. The specific working method is as follows:

[0112] 1. Obtain time series analysis prediction value

[0113] Time series analysis: First, use the improved autoregressive moving average model (as shown in formula (1)) to predict the trend term (T_t) of cooling demand.

[0114] Trend term prediction: According to historical data and model parameters, calculate the cooling demand trend value at each time point.

[0115] 2. Obtain multiple regression prediction value

[0116] Multiple regression analysis: Then, use the multiple regression model (as shown in formula (2)) to predict the cooling demand.

[0117] Regression model prediction: Input each factor (environmental temperature, humidity, door opening times, etc.) that affects cooling demand into the model, and calculate the cooling demand prediction value at each time point.

[0118] 3. Determine the weight coefficient

[0119] Weight selection: According to the actual situation and model performance, determine the weight coefficient (α) of the trend term and the weight and coefficient (β) of the regression phase. The selection of weight coefficient is usually based on the accuracy of model prediction and the importance of prediction results.

[0120] Weight range: Ensure that the weight coefficients satisfy the condition 0≤α≤1 and α+β=1, which means that the sum of the weights of the trend term and the regression pair is 1.

[0121] 4. Calculate the final cold demand prediction value

[0122] Weighted average: Use formula (3) to weight the time series analysis prediction value (T_t) and the multiple regression prediction value (Y_t) to obtain the final cold demand prediction value (Y'_t). Apply the calculated weight coefficients α and β to the above formula to calculate the final prediction value at each time point. By this method, the advantages of time series analysis and multiple regression can be combined to provide a more comprehensive and accurate cold demand prediction.

[0123] Step three, distribute control instructions to drive execution through dynamic refrigeration adjustment module

[0124] The database unit converts the prepared refrigeration adjustment strategy into specific control instructions and sends them to the control execution unit; after receiving the instructions, the control execution unit adjusts the working state of the compressor through the compressor controller, adjusts the fan speed through the fan controller, and controls the electromagnetic valve action through the electromagnetic valve controller, to jointly realize dynamic adjustment of the refrigerator refrigeration system; the dynamic refrigeration adjustment algorithm adjusts the compressor speed and fan power, and the improved compressor speed adjustment formula is:

[0125]

[0126] In formula (4), N(t) represents the target speed of the compressor at time t, N 基础 represents the basic speed of the compressor, e(t) represents the error signal, the deviation between the current temperature and the target temperature, u(t) represents the output signal, K p represents the proportional gain coefficient, K i represents the integral gain coefficient, and K d represents the differential gain coefficient.

[0127] The improved fan power adjustment output function is:

[0128]

[0129] In formula (5), P(t) represents the target speed of the compressor at time t, P 基础 represents the basic speed of the compressor, e(t) represents the error signal, the deviation between the current temperature and the target temperature, K p represents the proportional gain coefficient, K i represents the integral gain coefficient, and K d represents the differential gain coefficient.

[0130] In specific embodiments, the dynamic refrigeration regulation module adopts an advanced three-closed-loop control architecture in the compressor speed regulation system, including temperature loop, current loop and speed loop; intelligent PID regulation is realized. The system will collect temperature sensor data in real time, accurately calculate error signals, and adjust parameters online through fuzzy self-adaptive algorithm. Its working state is divided into standby, normal and emergency modes, standby mode maintains the basic speed, and the speed is 1200 rpm; the normal mode dynamically adjusts the speed in the range of 800-2000 rpm according to the real-time error; when the temperature deviation exceeds 2℃, the emergency mode is entered and the maximum refrigeration power is started. The fan power regulation system adopts variable air volume (VAV) control technology to adjust the power. The speed of the brushless DC motor is controlled by PWM signal, and the adjustment range can reach 0-100%, and the air volume sensor is integrated to realize closed-loop control. Its working state has low speed, medium speed and high speed modes, the power is ≤50% in low speed mode at night under low load, the power is 50-80% in medium speed mode under normal operation, and the power is greater than 80% in high speed mode during frequent opening period. The electromagnetic valve control system drives the expansion valve with a stepper motor, with a resolution of 0.01 mm, dynamically adjusts the opening degree based on pressure sensor data, and forms a linkage control strategy with the compressor speed to ensure the stable operation of the refrigeration system. In terms of control, multi-parameter fusion control is adopted, integrating temperature, pressure, humidity and door opening frequency parameters, and the adjustment accuracy is improved by 40% compared with the traditional single temperature control method. In terms of predictability, based on the improved time series algorithm, the cold demand can be predicted 15 minutes in advance, and the response speed is 60% faster than the reactive control. In terms of energy efficiency optimization, the compressor adopts direct current frequency conversion technology, and the energy efficiency ratio reaches 3.8, while the traditional fixed frequency is only 2.5; the fan adopts EC motor, and the efficiency is improved by 35%, and the dynamic adjustment strategy makes the system COP value increase by 22%. In terms of reliability, the over-temperature protection response time is less than 50ms, which is much lower than the industry standard of 200ms, and it also has the function of liquid knocking protection which traditional systems do not have.

[0131] Further, the working method of the intelligent defrosting control module is:

[0132] The frost thickness sensor collects data in the refrigerator and transmits the data to the sensing data fusion module; the sensing data fusion module determines defrosting, sends a defrosting instruction to the control execution unit, the control execution unit receives the defrosting instruction, the heating control module starts the heating device of the evaporator, heats the evaporator according to the set heating power, and melts the frost; the compressor and fan control module adjusts the operating state of the compressor and fan according to the defrosting demand; the drain control module opens the drain valve or starts the drain pump after the frost is melted to drain the melted water out of the refrigerator.

[0133] In specific embodiments, the intelligent defrosting control module works based on accurate collection and processing of multi-sensor data. The frost thickness sensor continuously collects data inside the refrigerator and transmits it to the sensor data fusion module. This module uses advanced algorithms to analyze and fuse multi-source data, accurately determining whether defrosting is needed. Once defrosting is determined, the sensor data fusion module quickly sends defrosting instructions to the control execution unit. After receiving the instructions, the control execution unit works collaboratively with each sub-module. The heating control module immediately starts the evaporator's heating device, heating the evaporator according to pre-set parameters to promote rapid melting of the frost layer. The compressor and fan control module adjusts the operation state of the compressor and fan according to the defrosting requirements to ensure efficient and stable defrosting process. The drainage control module opens the drainage valve or starts the drainage pump in a timely manner after the frost melts, successfully draining the melted water from the refrigerator, avoiding water accumulation affecting the performance of the refrigerator. The use of multi-sensor fusion technology can more comprehensively and accurately grasp the actual conditions inside the refrigerator, avoiding the limitations of single sensor judgment, making defrosting decisions scientific and reasonable. The heating power control strategy can accurately adjust the heating power according to different frost thickness and environmental conditions, ensuring defrosting effect and effectively reducing energy consumption. The intelligent compressor and fan cooperative control mode can optimize the overall operation efficiency of the refrigerator during the defrosting process, reducing unnecessary energy loss. In terms of technical effects, this module significantly improves the accuracy and timeliness of defrosting. Through accurate sensor data collection and analysis, defrosting can be started in the early stages of frost formation, avoiding the impact of thick frost on refrigeration effect. The efficient defrosting process significantly improves the refrigeration performance of the refrigerator, significantly reduces temperature fluctuations, and ensures the storage quality of items inside the refrigerator. At the same time, energy saving effect is remarkable, compared with traditional defrosting method, energy consumption can be reduced by about 30%-40%. In addition, the stability and reliability of this module are extremely high, with modular design, easy maintenance and upgrading, which can effectively prolong the service life of the refrigerator and reduce maintenance costs.

[0134] Further, the working method of the monitoring management module is:

[0135] The monitoring management module monitors the running state of the refrigerator in real time. The communication module receives various data from the refrigerator intelligent control system, the information collection module stores the data, the data sorting module cleans, classifies and analyzes the data, and the remote monitoring system provides an intuitive interface for users to view the running state of the refrigerator and can perform corresponding control operations according to requirements.

[0136] In a specific embodiment, the monitoring and management module monitors the freezer's operating status in real time. The communication module connects to the freezer's intelligent control system via the internet, receiving operating data such as temperature, humidity, and pressure collected from various sensors within the freezer, ensuring stable and accurate data transmission. The database unit receives data from the communication module and stores it according to a preset format and structure, providing a data foundation for subsequent operations. The data processing module cleans, classifies, and analyzes the stored data, removing duplicates and errors, handling missing values, and classifying the data by time and type to extract data value. The remote monitoring system provides users with an intuitive interface that displays the freezer's operating status in real time. Users can view data and remotely control the freezer as needed, such as adjusting the temperature and switching the refrigeration system on and off. This module performs stably under different operating conditions. Under normal conditions, each module operates in an orderly manner according to a predetermined process, continuously monitoring and managing the freezer; when abnormal data occurs, such as the temperature exceeding the range, an alarm will be issued promptly. The monitoring and management module can monitor the freezer's operating status in real time and comprehensively, with timely and comprehensive information collection and a scientifically sound processing procedure. In-depth data analysis enables precise identification of potential problems, early warning, and prevention of escalation of faults. Remote control functionality facilitates user operation and improves management efficiency. From a technical perspective, this module significantly enhances the intelligence and automation of freezer management, reduces manual inspection costs and labor intensity, improves the timeliness of fault handling, ensures stable freezer operation, extends freezer lifespan, and optimizes energy consumption, thereby reducing operating costs.

[0137] Furthermore, the diagnostic correction module operates as follows:

[0138] The collected data is preprocessed and then input into the fault diagnosis model. The fault diagnosis model compares and analyzes the fault patterns in the knowledge base to determine whether the freezer has a fault, as well as the type and severity of the fault. Based on the fault diagnosis results, the control module will determine whether the fault can be resolved through self-repair measures.

[0139] In a specific embodiment, the diagnostic and correction module enables intelligent diagnosis and repair of refrigerator malfunctions. A multi-channel ADC acquisition module performs 16-bit high-precision sampling of refrigerator operating data, which is then preprocessed with digital filtering to form a standardized data stream. The fault diagnosis module, based on a preset fault characteristic threshold library, uses a fuzzy logic inference algorithm to compare abnormal data patterns in real time. It determines the degree of evaporator frosting by the temperature gradient change rate and identifies compressor seizure faults using current fluctuation curves. When a fault signal is detected, the system automatically triggers a three-level response mechanism: Level 1 faults include abnormal door switch operation, which initiates a local self-repair program; Level 2 faults include abnormal fan speed, which is compensated by dynamically adjusting PID parameters; and Level 3 faults include compressor overload, which immediately triggers shutdown protection and pushes an alarm to the maintenance terminal.

[0140] Further, the working method of the energy-saving module is:

[0141] After real-time collection of the temperature, humidity, pressure, current, voltage and door switch state data of the refrigerator, the data processing module pre-processes the received data and inputs it into the energy-saving strategy model; the execution and control part controls the compressor, fan and valve components of the refrigerator through the compressor control module, fan control module and valve control module according to the energy-saving strategy, and adjusts the operating parameters of the refrigerator.

[0142] In specific embodiments, the energy-saving module collects the temperature, humidity, pressure, current, voltage and door switch state of the refrigerator in real time through sensors. After the data is collected in the data processing module, it will be pre-processed to remove noise, fill in missing values and unify the data format. The processed data is input into the energy-saving strategy model. According to the preset energy-saving algorithm and rules, combined with historical data and real-time working conditions, the optimal energy-saving strategy is analyzed. The execution and control part will control the compressor, fan and valve according to the energy-saving strategy with the help of the compressor control module, fan control module and valve control module. In different working conditions, the module shows flexibility. When the refrigerator is in a low load state, such as at night or with less goods, the energy-saving strategy model will reduce the operating frequency of the compressor, slow down the fan speed and adjust the valve opening to reduce unnecessary energy consumption; while in a high load state, such as during the day with high traffic and frequent door opening and closing, the operating parameters of each component will be dynamically optimized to balance energy saving and performance while ensuring the refrigeration effect. Compared with the traditional refrigerator control method, the energy-saving module has obvious advantages. The energy-saving module can dynamically adjust the operating parameters according to the real-time working conditions, avoiding the energy waste caused by the fixed mode of operation in the traditional way. Through precise data analysis and intelligent control, the module significantly improves the energy utilization efficiency. In terms of technical effect, through actual test, the energy-saving module can reduce the energy consumption of the refrigerator by 20%-30%, effectively reducing the operating cost. At the same time, it can better maintain the temperature stability in the refrigerator, reduce the fluctuation range and improve the preservation quality of the goods.

[0143] Although the specific embodiments of the present application are described above, those skilled in the art should understand that these specific embodiments are only illustrative, and those skilled in the art can make various omissions, substitutions and changes to the details of the above method and system without departing from the principles and essence of the present application. For example, combining the above method steps, performing substantially the same function in substantially the same way to achieve substantially the same result according to the same method is within the scope of the present application. Therefore, the scope of the present application is only limited by the appended claims.

Claims

1. An AI technology-based energy-saving mode intelligent commercial refrigerator system, characterized in that: The intelligent commercial refrigerator table is provided with an AI intelligent control module, a multi-dimensional sensor network module, a dynamic refrigeration adjustment module, an intelligent defrosting control module, a monitoring management module, a diagnosis correction module and an energy saving module. The AI intelligent control module controls and analyzes multi-dimensional data in the operation process of the refrigerator in real time, and comprises a central processor module, a data acquisition module connected with the central processor module, an improved sparrow algorithm module, an over-temperature calculation module and a data output interface. The multi-dimensional sensor network module is used for sensing refrigerator temperature, humidity, door opening times, article pressure and frost layer data, and is provided with temperature sensors, humidity sensors, weight sensors, pressure sensors and light sensors arranged inside and outside the refrigerator. The dynamic refrigeration adjustment module adjusts refrigeration intensity and operation state in real time according to feedback data of the AI intelligent control module, and comprises a database unit transmitted by the AI intelligent control module, a data processing module, an improved time series analysis module, a cold quantity demand prediction algorithm module and an execution mechanism. The intelligent defrosting control module performs intelligent defrosting by monitoring the frosting condition of the evaporator in real time through sensors, and comprises a frost layer sensor, a sensor data fusion module, a control execution unit, a heating control module and a drainage control module. The monitoring management module can view the operation state of the refrigerator in real time through the internet connection of the refrigerator intelligent control system, and comprises a communication module and an information collection module, a data arrangement module and a remote monitoring system connected with the communication module. The diagnosis correction module predicts and diagnoses potential faults, and comprises a fault diagnosis model and a self-repair execution unit. The energy saving module performs energy saving and consumption reduction of the refrigerator according to the data transmitted by the AI intelligent control module. The output ends of the multi-dimensional sensor network module, the dynamic refrigeration adjustment module, the intelligent defrosting control module, the monitoring management module and the energy saving module are connected with the input end of the AI intelligent control module, the output end of the AI intelligent control module is connected with the input ends of the dynamic refrigeration adjustment module, the intelligent defrosting control module, the monitoring management module, the multi-dimensional sensor network module and the energy saving module, the output ends of the monitoring management module and the intelligent defrosting control module are connected with the input end of the diagnosis correction module, the output end of the energy saving module is connected with the input end of the dynamic refrigeration adjustment module, and the output end of the diagnosis module is connected with the input end of the AI intelligent control module. 2.The AI technology-based energy-saving mode intelligent commercial refrigerator of claim 1, wherein: The working method of the AI intelligent control module is as follows: Step one, obtain refrigerator working state data information; The data acquisition module dynamically acquires data of temperature, humidity, pressure, current, voltage, door opening state and light intensity of the refrigerator, and transmits the collected and processed data to the data processing and analysis unit; Step two, calculate the optimal temperature through the improved sparrow algorithm module; The improved sparrow algorithm module takes the fan speed and the refrigeration valve opening as optimization variables, optimizes the fan speed and the refrigeration valve opening parameters to minimize the temperature fluctuation, and accurately controls the temperature; Step three, the over-temperature calculation module The over-temperature calculation module judges whether the equipment temperature exceeds the safe or normal operation range and performs statistics and analysis on the over-temperature condition, and the working method is as follows: According to the working requirements and safety standards of the equipment, a temperature threshold is set in advance, and the over-temperature threshold of the refrigeration area is set; the processing unit will receive the temperature data collected by the temperature sensor in real time, and compare it with the set threshold; if the current temperature exceeds the threshold, it is determined that the equipment is in an over-temperature state; Otherwise, it is in a normal temperature state; Step four, through the data output interface The integrated data is sent to the compressor in the energy-saving mode through the data output module. 3.The AI technology-based energy-saving mode intelligent commercial refrigerator of claim 1, wherein: The temperature sensor monitors the temperature of different areas inside the refrigerator in real time to ensure that the temperature is in an appropriate range; the humidity sensor detects the humidity in the refrigerator; the light sensor is installed inside the refrigerator, and when the door is opened, the light sensor detects the change of light to determine the opening state of the door; the pressure sensor is installed on the pipeline of the refrigeration system to monitor the pressure change of the refrigerant; the weight sensor is installed on the shelf of the refrigerator to monitor the weight change of the goods in the refrigerator in real time. 4.The AI technology-based energy-saving mode intelligent commercial refrigerator of claim 1, wherein: The working method of the dynamic refrigeration adjustment module is as follows: Step one, store the refrigerator data information through the database unit; The sensor continuously collects the data of the temperature, humidity, pressure, and weight of the goods inside the refrigerator and the external environment temperature of the refrigerator, and transmits them to the database unit; the database unit cleans, arranges and analyzes the collected data, identifies and processes the abnormal values and noises in the data; Step two, analyze and calculate the input data volume through the improved time series analysis module; Using the improved time series analysis combined with the method of multiple regression, the time series analysis part, through the improved autoregressive moving average model to predict the trend of cold demand Q 趋势 : In Equation (1), Q 趋势 represents a trend item of the cooling demand, Q 历史 represents historical cooling demand data, p represents an autoregressive order, q represents a moving average order, r represents a cooling growth order, represents an autoregressive order corresponding coefficient, θ j represents a moving average order corresponding coefficient, δ z represents a cooling growth amount, and ∈t-j represents a past error item; The multiple regression part establishes an improved refrigeration demand formula: In Equation (2), Q 需求 (t)) represents the amount of coldness required, β0represents a constant term, β 1- β3is the regression coefficient of each factor, T 环境 represents the ambient temperature, H 湿度 represents the humidity, N 开门次数 represents the number of times the door is opened, β i represents the sum of the regression coefficients, and ρ i represents the total number of goods; The improved final refrigeration demand prediction value is the trend item and the multiple regression weighted sum formula: Q 预测 (t) = aQ 趋势 (t) + βQ 需求 (t) + (1 - aβ)Q 趋势 (t)Q 需求 (t)(3) In Equation (3), Q 预测 (t) represents the final cooling demand prediction value, Q 趋势 (t) represents the cooling demand trend item based on time series analysis, Q 需求 (t) represents the cooling demand prediction value based on multiple regression, and α represents the weight coefficient of the trend item, where 0≤α≤1, and β represents the weight and coefficient of the regression phase. Step three, distribute the control instructions through the dynamic refrigeration adjustment module; The database unit converts the prepared refrigeration adjustment strategy into control instructions and sends them to the control unit; after receiving the instructions, the control unit adjusts the working state of the compressor, the fan speed of the fan controller, and the electromagnetic valve controller to realize the dynamic adjustment of the refrigerator refrigeration system; the dynamic refrigeration adjustment algorithm adjusts the compressor speed and the fan power, and the improved compressor speed adjustment formula is as follows: In Equation (4), N(t) denotes a target rotational speed of the compressor at time t, N 基础 denotes a base rotational speed of the compressor, e(t) denotes an error signal, a deviation of a current temperature from a target temperature, K p denotes a proportional gain coefficient, K i denotes an integral gain coefficient, and K d denotes a differential gain coefficient. The improved fan power adjustment output function is as follows: In Equation (5), P(t) denotes a target rotational speed of the compressor at time t, P 基础 denotes a base rotational speed of the compressor, e(t) denotes an error signal, a deviation of a current temperature from a target temperature, K p denotes a proportional gain coefficient, K i denotes an integral gain coefficient, and K d denotes a differential gain coefficient. 5.The AI technology-based energy-saving mode intelligent commercial refrigerator according to claim 1, characterized in that: The working method of the intelligent defrosting control module is as follows: The frost thickness sensor collects data in the refrigerator and transmits the data to the sensing data fusion module; the sensing data fusion module determines after defrosting, sends a defrosting instruction to the control execution unit, the control execution unit receives the defrosting instruction, the heating control module starts the heating device of the evaporator, heats the evaporator according to the set heating power, and melts the frost layer; the compressor and the fan control module adjust the running state of the compressor and the fan according to the defrosting demand; the drain control module opens the drain valve or starts the drain pump after the frost layer is melted to drain the melted water out of the refrigerator. 6.The AI technology-based energy-saving mode intelligent commercial refrigerator according to claim 1, characterized in that: The working method of the monitoring management module is: The monitoring management module monitors the running state of the refrigerator in real time; the communication module receives various data from the refrigerator intelligent control system, the information collection module stores the data, the data arrangement module cleans, classifies and analyzes the data, and the remote monitoring system provides an intuitive interface for users to view the running state of the refrigerator and can perform corresponding control operations according to the requirements. 7.The AI technology-based energy-saving mode intelligent commercial refrigerator according to claim 1, wherein: The working method of the diagnosis correction module is: The collected data is preprocessed, and the processed data is input into the fault diagnosis model; the fault diagnosis model compares and analyzes the fault mode in the knowledge base to determine whether the refrigerator has a fault, the type of the fault and the severity of the fault; According to the fault diagnosis result, the control module can judge whether the fault can be solved by self-repairing measures. 8.The AI technology-based energy-saving mode intelligent commercial refrigerator according to claim 1, wherein: The working method of the energy-saving module is: Real-time collection of temperature, humidity, pressure, current, voltage and door switch state data of the refrigerator, data processing, data arrangement module pre-processes the received data and inputs it into the energy-saving strategy model; the execution and control part controls the compressor, fan and valve components of the refrigerator according to the energy-saving strategy through the compressor control module, fan control module and valve control module, and adjusts the running parameters of the refrigerator.

Images