Refrigerated food Internet of Things temperature and humidity monitoring and control system

Abstract

The invention relates to the field of refrigerated food humidity monitoring, in particular to a refrigerated food Internet of Things temperature and humidity monitoring and control system which comprises a temperature and humidity sensor, a data acquisition module, a wireless communication module, a monitoring and early warning module, an intelligent control module and a system management module. According to the method, a network energy consumption optimization method is adopted to optimize energy consumption management and battery life, an RSA algorithm is adopted to select an optimal cluster head node to save energy, and an exploration stage and a development stage in the RSA algorithm are improved to ensure that tasks can be effectively executed within a relatively long time; according to the method, the parameters in the MPC strategy are optimized by adopting the automatically adjusted cost function, so that the MPC strategy can automatically adjust the working state of the refrigeration equipment according to the monitoring result, and intelligent control over the temperature and the humidity in the refrigeration environment is achieved.

Description

Technical Field

[0001] The present invention relates to the field of refrigerated food humidity monitoring, and in particular to a refrigerated food Internet of Things temperature and humidity monitoring and control system. Background Art

[0002] The existing refrigerated food temperature and humidity management system uses advanced sensor technology and wireless communication methods to achieve real-time monitoring of temperature and humidity changes in the refrigerated environment, thereby helping staff to more effectively control and maintain the refrigerated state. However, there are also some problems with the existing system. In terms of control, although it can achieve automatic monitoring and alarm, it is impossible to achieve accurate temperature and humidity control under strict temperature and humidity requirements, especially in complex and changing environments; secondly, energy consumption in wireless sensor networks is also an issue that cannot be ignored, especially for those wireless sensor nodes that rely on battery power. Their energy consumption management and battery life have become key factors for the long-term stable operation of the system; therefore, how to optimize system design, reduce power consumption, and improve control accuracy is one of the main challenges facing the current system. Summary of the invention

[0003] In view of the above situation, the present invention provides a refrigerated food Internet of Things temperature and humidity monitoring and control system. In view of the problems of energy consumption management and battery life in wireless sensor networks, a network energy consumption optimization method is used to optimize energy consumption management and battery life. The method adopts RSA (Reptile Search algorithm) algorithm to select the best cluster head node to save energy, and improves the exploration stage and development stage in the RSA algorithm. In the exploration stage, the algorithm pays more attention to finding new candidate nodes with sufficient energy as potential cluster heads; while in the development stage, it focuses on optimizing the energy utilization efficiency of existing cluster head nodes to ensure that they can effectively perform tasks for a long time; in view of the problem that in some cases, although automatic monitoring and alarm can be achieved in control, accurate temperature and humidity control may not be achieved, an automatically adjusted cost function is used to optimize the parameters in the MPC (Model Predictive Control) strategy, so that the MPC strategy can automatically adjust the working state of the refrigeration equipment according to the monitoring results, thereby realizing intelligent control of the temperature and humidity in the refrigeration environment.

[0004] The present invention provides a refrigerated food Internet of Things temperature and humidity monitoring and control system, including a temperature and humidity sensor, a data acquisition module, a wireless communication module, a monitoring and early warning module, an intelligent control module and a system management module;

[0005] The temperature and humidity sensors are used to collect and monitor the temperature changes and air humidity levels in the refrigerated environment and the refrigerated equipment, generate sensor data, and transmit them to the data acquisition module;

[0006] The data acquisition module is responsible for collecting and integrating sensor data and converting the sensor data into digital signals for transmission;

[0007] The wireless communication module uses a wireless sensor network for data transmission, transmits digital signals from the data acquisition module to the monitoring and early warning module and the system management module, and uses a network energy consumption optimization method to optimize the energy consumption in the wireless sensor network to extend the battery life;

[0008] The monitoring and early warning module continuously monitors the temperature and humidity of the refrigerated environment, and sets different temperature and humidity ranges for different areas in the refrigerated environment. When the temperature and humidity deviate from the set range, a warning is issued in time, and a monitoring report is generated and transmitted to the intelligent control module;

[0009] The intelligent control module uses an automatically adjusted cost function to learn the MPC strategy, automatically adjusts the working state of the refrigeration equipment according to the monitoring report, realizes intelligent control of the temperature and humidity in the refrigeration environment, and generates control records;

[0010] The system management module provides a user interface for system configuration and management.

[0011] Furthermore, the key components of the wireless sensor network include sensor nodes, base stations, network protocols and applications;

[0012] The wireless communication module optimizes the energy consumption in the wireless sensor network using a network energy consumption optimization method, and the network energy consumption optimization method specifically includes the following steps:

[0013] Step S1: Node initialization, dividing the sensor nodes in the wireless sensor network into M clusters, each cluster contains a cluster head;

[0014] Step S2: Select cluster head nodes, calculate the average distance between cluster heads and sensor nodes, the average distance between cluster heads and base stations, and the total energy of cluster heads, and find the best cluster head node combination for data transmission;

[0015] Step S3: Cluster construction, for each sensor node, calculate the weight value of the weight function of the sensor node and each cluster head, and select the cluster head with the highest weight value to merge and generate a new cluster;

[0016] Step S4: Search algorithm, use the improved RSA algorithm to optimize steps S2 and S3, select the best cluster head, and perform data transmission;

[0017] Further, in step S2, a cluster head node is selected, which specifically includes the following steps:

[0018] Step S21: For each cluster head, calculate the average distance between the cluster head and the sensor node. The formula used is as follows: ;

[0019] in, represents the number of cluster heads, is the total number of sensor nodes, represents the distance between the sensor node and each cluster head, and is the index, Indicates Cluster heads, Indicates sensor nodes, represents the average distance between the cluster head and the sensor nodes;

[0020] Step S22: Calculate the average distance between the cluster head and each base station, using the following formula: ;

[0021] in, represents a base station, is the average distance between the cluster head and each base station;

[0022] Step S23: Minimize the distance between the cluster head and the sensor node and the distance between the base station and each cluster head. The formula used is as follows: ;

[0023] In the formula, It means minimizing the distance between cluster heads and sensor nodes and the distance between base stations and each cluster head;

[0024] Step S24: Maximize the total energy of the cluster head, construct a fitness function, and find the best cluster head node combination. The formula used is as follows: ; ;

[0025] in, Cluster head The current energy value, is the inverse of the total energy to be minimized, is the control factor, is the fitness function;

[0026] Further, in step S4, the improved RSA algorithm specifically includes the following steps:

[0027] Step S41: population initialization, randomly generating an initial population, setting the problem dimension and the maximum number of iterations;

[0028] Step S42: In the high-speed walking exploration phase of the population individuals, the exploration method is selected according to the number of iterations. When , the formula used is as follows: ;

[0029] in, Indicates The individual in The position of the problem dimension, Indicates the current iteration number, Represents the current The best position on the problem dimension, For the hunting operator, To control the parameters, is the reduction function, Represents the value of the chaotic map;

[0030] Step S43: The population individual ventral exploration phase, when When , the formula used is as follows: ;

[0031] in, The random solution Locations, is a random number between [1,N], is the evolution perception function;

[0032] Step S44: Population individual coordination development stage, when When , the formula used is as follows: ;

[0033] in, Indicates The individual in The percentage difference between the position on the question dimension and the average position;

[0034] Step S45: Population individual cooperative development stage, when When , the formula used is as follows: ;

[0035] in, It is a decimal value, the value is 0.0001.

[0036] Furthermore, the intelligent control module uses an automatically adjusted cost function to learn the MPC strategy, which specifically includes the following steps:

[0037] Step M1: Define the linear MPC problem. The linear MPC problem can be expressed as: ;

[0038] The constraints are: ; ; ;

[0039] in, and is the combination of state variables and control variables, represents the length of the time domain, is the index, is the transpose operation, and are the cost function and cost vector matrix respectively, represents the residual in the dynamics, is a matrix, Indicates that MPC is at time The state vector of and represents the MPC initial state vector, is a vector containing the upper limits of all inequality constraints, represents a matrix that defines inequality constraints on control inputs and states;

[0040] Step M2: Solve the linear MPC problem using a standard QP solver to obtain the optimal solution for the combination of state variables and control variables;

[0041] Step M3: Construct the gradient problem of linear MPC and construct an auxiliary problem for calculating the gradient, expressed as: ;

[0042] The constraints are: ; ; [ G k ] It t ̃ k =0 ;

[0043] in, is an auxiliary state-control combination, [ G k ] It Represented in the matrix So that the inequality A row of values ​​that are active in the optimal solution, is the vector of cost functions in the auxiliary problem, and is the initial state vector of the auxiliary problem, Indicates that the auxiliary problem is in time The state vector of

[0044] Step M4: Calculate the gradient of the linear MPC,

[0045] Step M5: Process active constraints, set constraint boundaries of control variables, and determine whether the control variables have reached the constraint boundaries. When the control variables have reached the constraint boundaries, the gradient of the control variables is zero.

[0046] Step M6: nonlinear MPC, for nonlinear MPC problems, the SQP method is used for approximate solution, and the SQP method generates and optimizes the nonlinear MPC problem in an iterative manner;

[0047] Step M7: Use the gradient of the linear MPC to calculate the gradient of the nonlinear MPC to obtain the target gradient.

[0048] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0049] (1) Aiming at the problems of energy management and battery life in wireless sensor networks, a network energy optimization method is adopted to optimize energy management and battery life. The core of this method is to save energy by optimizing the selection of cluster head nodes. The exploration phase and development phase of the existing RSA algorithm are improved. In the exploration phase, this method pays more attention to finding new candidate nodes with sufficient energy as potential cluster heads; while in the development phase, it focuses on optimizing the energy efficiency of existing cluster head nodes to ensure that they can effectively perform tasks over a longer period of time, which helps to balance the energy consumption in the entire network and can significantly extend the lifetime of the wireless sensor network, so that the network can continue to operate for a longer time.

[0050] (2) In terms of control, although automatic monitoring and alarm can be achieved, accurate temperature and humidity control may not be achieved in some cases. Therefore, an automatically adjusted cost function is used to optimize the parameters in the MPC strategy. This method continuously monitors the temperature and humidity in the cold storage environment and uses feedback information to dynamically adjust the weights in the cost function to achieve accurate control of the equipment's working status. As the monitoring data changes, the MPC algorithm will update its control strategy in real time to cope with different changes in external conditions, thereby improving the flexibility and response speed of the system and ensuring that the best refrigeration performance can be maintained under various working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a refrigerated food Internet of Things temperature and humidity monitoring and control system provided by the present invention;

[0052] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0054] Example 1, see Figure 1 , a refrigerated food Internet of Things temperature and humidity monitoring and control system provided by the present invention includes a temperature and humidity sensor, a data acquisition module, a wireless communication module, a monitoring and early warning module, an intelligent control module and a system management module;

[0055] The temperature and humidity sensors are used to collect and monitor the temperature changes and air humidity levels in the refrigerated environment and the refrigerated equipment, generate sensor data, and transmit them to the data acquisition module;

[0056] The data acquisition module is responsible for collecting and integrating sensor data and converting the sensor data into digital signals for transmission;

[0057] The wireless communication module uses a wireless sensor network for data transmission. The devices in the wireless sensor network include Wi-Fi, Bluetooth, LoRaWAN and NB-IoT. The digital signal is transmitted from the data acquisition module to the monitoring and early warning module and the system management module. The network energy consumption optimization method is used to optimize the energy consumption in the wireless sensor network to extend the battery life.

[0058] The monitoring and early warning module continuously monitors the temperature and humidity of the refrigerated environment, and sets different temperature and humidity ranges for different areas in the refrigerated environment. When the temperature and humidity deviate from the set range, a warning is issued in time, and a monitoring report is generated and transmitted to the intelligent control module;

[0059] The intelligent control module uses an automatically adjusted cost function to learn the MPC strategy, automatically adjusts the working state of the refrigeration equipment according to the monitoring report, realizes intelligent control of the temperature and humidity in the refrigeration environment, and generates control records;

[0060] The system management module provides a user interface for system configuration and management. The user interface allows users to access the system through a web page or mobile application, set system parameters, including temperature range, relative humidity, alarm thresholds, control rules, and display monitoring report and control record information in graphical form to facilitate trend analysis and report generation.

[0061] Embodiment 2: This embodiment is based on the above embodiment, and the temperature and humidity sensors are thermistors, thermocouples and capacitive humidity sensors, so as to perform temperature and humidity monitoring;

[0062] The data acquisition module uses a microcontroller to process sensor data, performs preliminary data processing and logic judgment, uses an analog-to-digital converter to convert analog signals into digital signals, and uses a memory to temporarily store data until the wireless communication module transmits the stored data to the monitoring and early warning module and the system management module;

[0063] The intelligent control module is connected to the actuator of the refrigeration equipment using an actuator interface, so as to perform intelligent control using an MPC strategy.

[0064] Embodiment 3, based on the above embodiment, the key components of the wireless sensor network include sensor nodes, base stations, network protocols and applications;

[0065] The wireless communication module optimizes the energy consumption in the wireless sensor network using a network energy consumption optimization method, and the network energy consumption optimization method specifically includes the following steps:

[0066] Step S1: Node initialization, dividing the sensor nodes in the wireless sensor network into M clusters, each cluster contains a cluster head;

[0067] Step S2: Select cluster head nodes, calculate the average distance between cluster heads and sensor nodes, the average distance between cluster heads and base stations, and the total energy of cluster heads, and find the best cluster head node combination for data transmission;

[0068] Step S3: Cluster construction, for each sensor node, calculate the weight value of the weight function of the sensor node and each cluster head, and select the cluster head with the highest weight value to merge and generate a new cluster;

[0069] Step S4: Search algorithm, use the improved RSA algorithm to optimize steps S2 and S3, select the best cluster head, and perform data transmission.

[0070] Embodiment 4: This embodiment is based on the above embodiment. In step S2, a cluster head node is selected, which specifically includes the following steps:

[0071] Step S21: For each cluster head, calculate the average distance between the cluster head and the sensor node. The formula used is as follows: ;

[0072] in, represents the number of cluster heads, is the total number of sensor nodes, represents the distance between the sensor node and each cluster head, and is the index, Indicates Cluster heads, Indicates sensor nodes, represents the average distance between the cluster head and the sensor nodes;

[0073] Step S22: Calculate the average distance between the cluster head and each base station, using the following formula: ;

[0074] in, represents a base station, is the average distance between the cluster head and each base station;

[0075] Step S23: Minimize the distance between the cluster head and the sensor node and the distance between the base station and each cluster head. The formula used is as follows: ;

[0076] In the formula, It means minimizing the distance between cluster heads and sensor nodes and the distance between base stations and each cluster head;

[0077] Step S24: Maximize the total energy of the cluster head, construct a fitness function, and find the best cluster head node combination. The formula used is as follows: ; ;

[0078] in, Cluster head The current energy value, is the inverse of the total energy to be minimized, is the control factor, is the fitness function.

[0079] Embodiment 5, based on the above embodiment, in step S4, the improved RSA algorithm specifically includes the following steps:

[0080] Step S41: population initialization, randomly generating an initial population, setting the problem dimension and the maximum number of iterations;

[0081] Step S42: In the high-speed walking exploration phase of the population individuals, the exploration method is selected according to the number of iterations. When , the formula used is as follows: ;

[0082] in, Indicates The individual in The position of the problem dimension, Indicates the current iteration number, Represents the current The best position on the problem dimension, For the hunting operator, To control the parameters, is the reduction function, Represents the value of the chaotic map;

[0083] Step S43: The population individual ventral exploration phase, when When , the formula used is as follows: ;

[0084] in, The random solution Locations, is a random number between [1,N], is the evolution perception function;

[0085] Step S44: Population individual coordination development stage, when When , the formula used is as follows: ;

[0086] in, Indicates The individual in The percentage difference between the position on the question dimension and the average position;

[0087] Step S45: Population individual cooperative development stage, when When , the formula used is as follows: ;

[0088] in, It is a decimal value, the value is 0.0001.

[0089] Embodiment 6: This embodiment is based on the above embodiment. In step S42, The expression is: ;

[0090] In the formula The expression is: ;

[0091] In step S43, the formula The expression is: ;

[0092] In step S44, the formula The expression is: ;

[0093] in, is a sensitive parameter, equal to 0.1, Indicates The average position of individuals, and The location The upper and lower limits of and is a random number between [1,N], The random solution locations.

[0094] By performing the above operations, in view of the problems of energy consumption management and battery life in wireless sensor networks, a network energy consumption optimization method is used to optimize energy consumption management and battery life. The core of this method is to achieve the purpose of saving energy by optimizing the selection of cluster head nodes. The exploration stage and development stage in the existing RSA algorithm are improved. In the exploration stage, this method pays more attention to finding new and energy-sufficient candidate nodes as potential cluster heads; while in the development stage, it focuses on optimizing the energy utilization efficiency of existing cluster head nodes to ensure that they can effectively perform tasks over a longer period of time, which helps to balance the energy consumption in the entire network and can also significantly extend the lifetime of the wireless sensor network, so that the network can continue to run for a longer time.

[0095] Embodiment 7, based on the above embodiment, the intelligent control module adopts an automatically adjusted cost function to learn the MPC strategy, specifically comprising the following steps:

[0096] Step M1: Define the linear MPC problem. The linear MPC problem can be expressed as: ;

[0097] The constraints are: ; ; ;

[0098] in, and is the combination of state variables and control variables, represents the length of the time domain, is the index, is the transpose operation, and are the cost function and cost vector matrix respectively, represents the residual in the dynamics, is a matrix, Indicates that MPC is at time The state vector of and represents the MPC initial state vector, is a vector containing the upper limits of all inequality constraints, represents a matrix that defines inequality constraints on control inputs and states;

[0099] Step M2: Solve the linear MPC problem using a standard QP solver to obtain the optimal solution for the combination of state variables and control variables;

[0100] Step M3: Construct the gradient problem of linear MPC and construct an auxiliary problem for calculating the gradient, expressed as: ;

[0101] The constraints are: ; ; [ G k ] It t ̃ k =0 ;

[0102] in, is an auxiliary state-control combination, [ G k ] It Represented in the matrix So that the inequality A row of values ​​that are active in the optimal solution, is the vector of cost functions in the auxiliary problem, and is the initial state vector of the auxiliary problem, Indicates that the auxiliary problem is in time The state vector of

[0103] Step M4: Calculate the gradient of the linear MPC;

[0104] Step M5: Process active constraints, set constraint boundaries of control variables, and determine whether the control variables have reached the constraint boundaries. When the control variables have reached the constraint boundaries, the gradient of the control variables is zero.

[0105] Step M6: nonlinear MPC, for nonlinear MPC problems, the SQP method is used for approximate solution, and the SQP method generates and optimizes the nonlinear MPC problem in an iterative manner;

[0106] Step M7: Use the gradient of the linear MPC to calculate the gradient of the nonlinear MPC to obtain the target gradient.

[0107] Embodiment 8, based on the above embodiment, in step M7, the target gradient is obtained by using the chain rule, and the target gradient is , The parameter relationship is shown in the following table:

[0108] in, represents the first optimal control action, ∂ u g 1 ∂ [ C k ] ij Represents the cost function matrix Middle Line The change of column elements The impact of ∂ u g 1 ∂ [ c k ] ij Represents the cost vector Middle The change of the element The impact of ∂ u g 1 ∂ [ y init ] i Indicates the initial state Middle The change of the element The impact of Indicates that only Line The identity matrix with columns set to 1, express Middle elements are directly equal to No. Negative values ​​of elements, express The value of is the first optimal control action.

[0109] By performing the above operations, in terms of control, although automatic monitoring and alarm can be achieved, accurate temperature and humidity control may not be achieved in some cases. An automatically adjusted cost function is used to optimize the parameters in the MPC strategy. This method continuously monitors the temperature and humidity in the refrigerated environment, and uses feedback information to dynamically adjust the weights in the cost function to achieve accurate control of the working status of the equipment. As the monitoring data changes, the MPC algorithm will update its control strategy in real time to cope with different changes in external conditions, improve the flexibility and response speed of the system, and ensure that the best refrigeration performance can be maintained under various working conditions.

[0110] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0111] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A refrigerated food Internet of Things temperature and humidity monitoring and control system, characterized by: It includes temperature and humidity sensors, data acquisition modules, wireless communication modules, monitoring and early warning modules, intelligent control modules and system management modules; The temperature and humidity sensors are used to collect and monitor the temperature changes and air humidity levels in the refrigerated environment and the refrigerated equipment, generate sensor data, and transmit them to the data acquisition module; The data acquisition module is responsible for collecting and integrating sensor data and converting the sensor data into digital signals for transmission; The wireless communication module uses a wireless sensor network for data transmission, transmits digital signals from the data acquisition module to the monitoring and early warning module and the system management module, and uses a network energy consumption optimization method to optimize the energy consumption in the wireless sensor network to extend the battery life; The monitoring and early warning module continuously monitors the temperature and humidity of the refrigerated environment, and sets different temperature and humidity ranges for different areas in the refrigerated environment. When the temperature and humidity deviate from the set range, a warning is issued in time, and a monitoring report is generated and transmitted to the intelligent control module; The intelligent control module uses an automatically adjusted cost function to learn the MPC strategy, automatically adjusts the working state of the refrigeration equipment according to the monitoring report, realizes intelligent control of the temperature and humidity in the refrigeration environment, and generates control records; The system management module provides a user interface for system configuration and management.

2. A refrigerated food Internet of Things temperature and humidity monitoring and control system according to claim 1, characterized in that: The key components of the wireless sensor network include sensor nodes, base stations, network protocols and applications; The wireless communication module optimizes the energy consumption in the wireless sensor network using a network energy consumption optimization method, and the network energy consumption optimization method specifically includes the following steps: Step S1: Node initialization, dividing the sensor nodes in the wireless sensor network into M clusters, each cluster contains a cluster head; Step S2: Select cluster head nodes, calculate the average distance between cluster heads and sensor nodes, the average distance between cluster heads and base stations, and the total energy of cluster heads, and find the best cluster head node combination for data transmission; Step S3: Cluster construction, for each sensor node, calculate the weight value of the weight function of the sensor node and each cluster head, and select the cluster head with the highest weight value to merge and generate a new cluster; Step S4: Search algorithm, use the improved RSA algorithm to optimize steps S2 and S3, select the best cluster head, and perform data transmission.

3. A refrigerated food Internet of Things temperature and humidity monitoring and control system according to claim 2, characterized in that: In step S2, a cluster head node is selected, which specifically includes the following steps: Step S21: For each cluster head, calculate the average distance between the cluster head and the sensor node; Step S22: Calculate the average distance between the cluster head and each base station; Step S23: Minimize the distance between the cluster head and the sensor node and the distance between the base station and each cluster head; Step S24: Maximize the total energy of cluster heads, construct a fitness function, and find the best cluster head node combination.

4. A refrigerated food Internet of Things temperature and humidity monitoring and control system according to claim 2, characterized in that: In step S4, the improved RSA algorithm specifically includes the following steps: Step S41: population initialization, randomly generating an initial population, setting the problem dimension and the maximum number of iterations; Step S42: In the high-speed walking exploration phase of the population individuals, the exploration method is selected according to the number of iterations. When , the formula used is as follows: ； in, Indicates The individual in The position of the problem dimension, Indicates the current iteration number, Represents the current The best position on the problem dimension, For the hunting operator, To control the parameters, is the reduction function, Represents the value of the chaotic map; Step S43: The population individual ventral exploration phase, when When , the formula used is as follows: ； in, The random solution Locations, is a random number between [1,N], is the evolution perception function; Step S44: Population individual coordination development stage, when When , the formula used is as follows: ； in, Indicates The individual in The percentage difference between the position on the question dimension and the average position; Step S45: Population individual cooperative development stage, when When , the formula used is as follows: ； in, It is a decimal value, the value is 0.0001.

5. The refrigerated food Internet of Things temperature and humidity monitoring and control system according to claim 1, characterized in that: The intelligent control module uses an automatically adjusted cost function to learn the MPC strategy, which specifically includes the following steps: Step M1: Define the linear MPC problem; Step M2: Solve the linear MPC problem using a standard QP solver to obtain the optimal solution for the combination of state variables and control variables; Step M3: Construct the gradient problem of linear MPC and construct an auxiliary problem for calculating the gradient; Step M4: Calculate the gradient of the linear MPC, Step M5: Process active constraints, set constraint boundaries of control variables, and determine whether the control variables have reached the constraint boundaries. When the control variables have reached the constraint boundaries, the gradient of the control variables is zero. Step M6: nonlinear MPC, for nonlinear MPC problems, the SQP method is used for approximate solution, and the SQP method generates and optimizes the nonlinear MPC problem in an iterative manner; Step M7: Use the gradient of the linear MPC to calculate the gradient of the nonlinear MPC to obtain the target gradient.

Images