Adjustment method and system applied to terminal box environment controller, electronic equipment and storage medium

Abstract

The invention discloses an adjusting method and system applied to a terminal box environment controller, electronic equipment and a storage medium. The system comprises a multi-mode sensing module used for collecting multi-dimensional environment data of a terminal box; the fusion decision-making module is used for analyzing the collected data, calculating a risk index and providing a basis for adjustment decision-making; the self-adaptive dynamic adjusting module is used for switching a control mode and adjusting equipment operation parameters according to the risk index and the environment parameters; the cloud edge collaboration module is used for realizing local real-time decision and remote management functions; and the low-power consumption and reliability module is used for reducing the energy consumption operation of the system and guaranteeing the stable operation of the system under different power supply conditions. According to the invention, the environment of the terminal box can be dynamically adjusted according to changes of different environments, the accuracy of environment regulation and control is improved, and the stability of the environment of the terminal box is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment operation environment control, and specifically to an adjustment method and system, an electronic device, and a storage medium for a terminal box environment controller. Background Art

[0002] In the fields of electric power, communication, and industrial automation, the terminal box is a key hub for electrical equipment connection and distribution, and the stability of its internal operation environment is directly related to the reliability and safety of the entire system. An internal environment with appropriate temperature and humidity, no harmful gas interference, and stable mechanical conditions is an important prerequisite for ensuring the normal operation of various electrical components in the terminal box, extending the service life of the equipment, and reducing the failure rate. However, the traditional terminal box environment control technology has exposed many insurmountable limitations in practical applications.

[0003] In terms of environmental data collection, the traditional system relies heavily on a single temperature and humidity sensor, and the monitoring means are extremely limited. This single-dimensional monitoring method makes it impossible for the system to comprehensively perceive potential risk factors such as harmful gas accumulation, abnormal light, and mechanical vibration that may exist inside the terminal box. Due to the lack of this key information, it is difficult for the system to comprehensively and accurately evaluate the true operating environment of the terminal box, resulting in insufficient data support for subsequent control decisions and making it difficult to achieve precise and effective environmental regulation.

[0004] The traditional decision-making mechanism is mainly based on a simple threshold judgment logic. The system simply compares the collected temperature and humidity data with a pre-set fixed threshold, and once the data exceeds the threshold range, it triggers the corresponding control action. This simple decision-making mode completely ignores the complex correlations and comprehensive impacts between different environmental parameters and cannot scientifically and comprehensively evaluate environmental risks. In practical applications, multiple environmental factors often interact with each other, and a change in a single parameter may trigger a chain reaction, while the traditional threshold judgment mechanism cannot capture these potential risk changes, resulting in one-sidedness and lag in decision-making.

[0005] In the environmental adjustment link, the control mode of the traditional system is relatively rigid and lacks flexibility and adaptability. The operating parameters of the control devices (such as fans, heaters, dehumidifiers, etc.) are usually fixedly set or can only be switched between a limited number of gears, and cannot be adjusted in real time and precisely according to the dynamic changes of the actual environment. For example, when the environmental temperature and humidity change simultaneously, the traditional system cannot coordinately control the adjustment devices according to their comprehensive impacts, resulting in poor adjustment effects and unable to meet the strict requirements of the terminal box for environmental stability.

[0006] In terms of system collaboration, traditional terminal box environmental control systems mainly focus on local control, and the remote management function is relatively weak. Although some systems have certain remote data transmission and monitoring capabilities, there is a lack of an effective collaboration mechanism between local real-time decision-making and remote management. In the case of unstable or interrupted network, the remote management function often cannot function properly, resulting in a significant decrease in the reliability and management efficiency of the system. This disconnection between local and remote functions limits the adaptability and flexibility of the system in complex application scenarios.

[0007] In terms of power consumption management and reliability guarantee, traditional systems have obvious deficiencies. In terms of energy consumption, due to the lack of an effective power management strategy, there is often energy waste during the operation of the system, especially in application scenarios powered by batteries. This problem is particularly prominent, seriously affecting the battery life and operating cost of the system. In terms of reliability, traditional systems lack redundant design and fault tolerance mechanisms for key hardware devices. Once a key component fails, the entire system may become paralyzed and cannot ensure stable operation under various complex power supply conditions.

[0008] In summary, traditional terminal box environmental control technologies are difficult to meet the strict requirements for the operation stability, safety, and efficient management of equipment in today's digital and intelligent era. Therefore, it is necessary to provide a new technical solution to improve one or more problems existing in the above solutions. Summary of the Invention

[0009] To solve the deficiencies in the prior art, the present invention provides an adjustment method and system, an electronic device, and a storage medium for a terminal box environmental controller, which can dynamically adjust the terminal box environment according to changes in different environments, improve the accuracy of environmental regulation, and further enhance the stability of the terminal box environment.

[0010] The present invention adopts the following technical solutions.

[0011] The first aspect of the present invention provides an adjustment method for a terminal box environmental controller, including: Collect a variety of environmental data inside the terminal box, and preprocess the collected variety of environmental data, and trigger an alarm when any one of the environmental data exceeds the corresponding environmental parameter threshold; Input the historical environmental data of the box end and the environmental parameter threshold into a pre-trained decision tree model for classification judgment, and mark the classified normal environmental data and abnormal environmental data; Based on a preset loss function, input the classified normal environmental data and abnormal environmental data and their corresponding marks into a pre-constructed risk index calculation model for training to obtain a trained risk index calculation model; Perform risk assessment based on the pre-processed multiple environmental data inputs using a trained risk index calculation model to obtain the risk index value; Determine whether the risk index value exceeds the risk threshold. If it exceeds, determine the control mode based on the environmental parameter threshold and the risk index value, and adjust the operating parameters of the equipment in the terminal box according to the determined control mode.

[0012] Optionally, construct a risk index calculation model based on the linear regression algorithm, and initialize the weights of each parameter environment according to the probability of each environmental data under each environmental parameter when constructing the risk index calculation model.

[0013] Optionally, initialize the weights of each parameter environment according to the probability of each environmental data under each environmental parameter, including: Calculate the probability of each environmental data under each environmental parameter; Calculate the information disorder degree value of each environmental parameter according to the probabilities of each environmental data; Calculate the information utility value of each environmental parameter based on the information disorder degree value of each environmental parameter; Initialize the weights of each environmental parameter in the risk index calculation model according to the information utility value of each environmental parameter.

[0014] Optionally, initialize the weights of each environmental parameter in the risk index calculation model according to the following formula based on the information utility value of each environmental parameter: , where, represents the weight of the j-th environmental parameter, represents the information utility value of the j-th environmental parameter, represents the number of environmental parameters.

[0015] Optionally, calculate the preset loss function according to the following formula: , where, represents the loss value, represents the predicted value of the risk index value, represents the true value of the risk index, represents the regularization coefficient, represents the number of environmental data samples, represents the number of environmental parameters, represents the weight of the j-th environmental parameter in the risk index calculation model, represents the weight of the i-th environmental data sample.

[0016] Optionally, determine the control mode based on the environmental parameter threshold and the risk index value, including: Determine the control mode switching condition according to each environmental parameter threshold and risk index value; Using the fuzzy logic algorithm, taking the risk index value and multiple environmental data as inputs, and combining with the control mode switching condition for fuzzy inference to obtain the comprehensive evaluation result of the environmental state; Determine the control mode to be switched according to the comprehensive evaluation result.

[0017] Optionally, the method further includes: Receive a remote control instruction and adjust the operating parameters of the equipment in the terminal box according to the remote control instruction.

[0018] Optionally, the method further includes: Based on the reinforcement learning algorithm, dynamically adjust the supply voltage and current of each module in the terminal box controller according to the obtained power data, environmental data and equipment load of the terminal box controller.

[0019] The second aspect of the present invention provides an adjustment system applied to a terminal box environment controller, and the system includes: A multimodal environment perception module for collecting various environmental data in the terminal box; A cloud-edge collaboration module, including an edge computing unit, for preprocessing various environmental data collected by the multimodal environment perception module, triggering an alarm when any environmental data exceeds the corresponding environmental parameter threshold, and inputting the historical environmental data and environmental parameter threshold of the box end into a pre-trained decision tree model for classification judgment, and marking the classified normal environmental data and abnormal environmental data; A fusion decision module for training the classified normal environmental data, abnormal environmental data and their corresponding marks sent by the edge computing unit into a pre-constructed risk index calculation model based on a preset loss function to obtain a trained risk index calculation model; and inputting the various environmental data preprocessed by the edge computing unit into the trained risk index calculation model for risk assessment to obtain a risk index value; judging whether the risk index value exceeds the risk threshold, and in the case of exceeding, sending an adjustment command to the adaptive dynamic adjustment module; An adaptive dynamic adjustment module, including a control module switching unit and an equipment parameter adjustment unit. The control module switching unit is used to determine the control mode according to the environmental parameter threshold and risk index value after receiving the adjustment command sent by the fusion decision module, and the equipment parameter adjustment unit is used to adjust the operating parameters of the equipment in the terminal box according to the determined control mode.

[0020] Optionally, the cloud-edge collaboration module further includes a cloud server unit, which is configured to receive and store the preprocessed data uploaded by the edge computing unit, classify the preprocessed data using a clustering analysis algorithm to obtain the device operation mode rules, and display the device operation mode rules to the user. At the same time, it receives the remote control instructions from the user for the terminal box environment controller. The remote control instructions include the setting of the terminal box environment parameters, and the operation parameters of the devices in the terminal box are adjusted according to the remote control instructions.

[0021] Optionally, the adjustment system further includes a low-power and reliability module, which includes an adaptive power regulation unit, a hardware device redundancy switching unit, and a fault intelligent error tolerance processing unit; among them, The adaptive power regulation unit is configured to dynamically adjust the supply voltage and current of each module in the terminal box controller based on a reinforcement learning algorithm according to the obtained power data, environmental data, and device load of the terminal box controller; The terminal box environment controller includes a main hardware and a backup hardware. The hardware device redundancy switching unit is configured to perform real-time synchronous status monitoring and data backup on the main hardware and the backup hardware, and automatically switch to the backup hardware when the main hardware fails; The fault intelligent error tolerance processing unit is configured to use a cyclic redundancy check algorithm to check the data collected by the multi-modal environment perception module and transmitted to the edge computing unit during the data transmission process. When it detects that there is a transmission error in the data, it generates a retransmission instruction and requests the multi-modal environment perception module to resend the data through the cloud server unit. A third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is loaded into the processor, it implements the above-mentioned adjustment method applied to the terminal box environment controller.

[0022] A fourth aspect of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned adjustment method applied to the terminal box environment controller.

[0023] Compared with the prior art, the beneficial effects of the present invention at least include: 1. Comprehensive environmental perception: The multi-modal environment perception module of the present invention integrates multiple sensor units such as temperature and humidity, gas, light, and vibration. Compared with traditional single temperature and humidity monitoring, it can collect multi-dimensional environmental data in the terminal box comprehensively and with high precision. This enables the system to have a more comprehensive and accurate understanding of the real environmental conditions of the terminal box, provides a rich and reliable data basis for subsequent decision-making and adjustment, and effectively avoids equipment operation risks caused by missing environmental information.

[0024] 2. Intelligent Decision-making: The fusion decision-making module abandons the traditional simple threshold judgment method through data preprocessing, in-depth data analysis, and scientific risk index calculation. It can not only uncover the potential relationships and patterns among different environmental parameters but also accurately calculate the risk index based on the impact degree of each parameter on the environment. This provides a scientific, quantitative, and comprehensive basis for adjustment decisions, greatly enhancing the accuracy and foresight of decisions, enabling the system to anticipate and respond to potential environmental risks in advance.

[0025] 3. Adaptive Dynamic Adjustment: The adaptive dynamic adjustment module can intelligently switch among multiple control modes according to the risk index and real-time environmental parameters and precisely adjust the operating parameters of the adjustment equipment dynamically. Compared with the traditional adjustment method with fixed modes and parameters, the present invention can better adapt to the complex and changeable environmental requirements of the terminal box, achieve more accurate and efficient environmental adjustment, and ensure that the inside of the terminal box always maintains a stable environment suitable for the operation of the equipment.

[0026] 4. Management Mode Collaboration: The cloud-edge collaboration module realizes the efficient collaboration between local real-time decision-making and remote management. The edge computing unit ensures the timeliness of local data processing and real-time decision-making, enabling the system to operate normally even in case of network anomalies; the cloud server unit provides powerful remote data storage, analysis, and management functions, facilitating users to monitor and manage the terminal box anytime and anywhere. This collaboration mode significantly improves the management efficiency and convenience of the system, breaking the limitation of the traditional system's separation between local and remote functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for description in the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them: Figure 1 It is a schematic flowchart of an adjustment method applied to an environmental controller of a terminal box provided by an embodiment of the present invention; Figure 2 It is a schematic architecture diagram of an adjustment system applied to an environmental controller of a terminal box provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0029] Combined Figure 1 As shown, Embodiment 1 of the present invention provides an adjustment method applied to a terminal box environment controller, including the following: S1. Collect a variety of environmental data in the terminal box, and preprocess the collected variety of environmental data. When any one of the environmental data exceeds the corresponding environmental parameter threshold, an alarm is triggered.

[0030] The multi-modal environmental perception module uses a variety of sensors to collect the environmental data of the terminal box under different working conditions, and measures the temperature and humidity, harmful gas concentration, light intensity, vibration conditions, smoke state, water seepage detection point state, etc. in the terminal box respectively, to obtain environmental data in real time; when the environmental parameter exceeds the threshold, abnormal data is transmitted and an alarm is triggered to prompt the operation and maintenance personnel to handle it.

[0031] The environmental parameter thresholds include temperature threshold, humidity threshold, maximum temperature rise, smoke threshold, harmful gas concentration threshold, light intensity threshold, vibration threshold, etc.

[0032] It can be understood that when any one of the environmental data exceeds the corresponding environmental parameter threshold, an alarm is triggered, which can be an audible and visual alarm, and different audible and visual alarms can correspond to different environmental parameters. In this way, the inspection personnel can be reminded to operate quickly, such as starting and stopping the dehumidifier, exhaust fan, heater, air conditioner, fire extinguishing device, etc.

[0033] The preprocessing of the variety of environmental data collected in S1 includes: After the multi-modal environmental perception module transmits the variety of environmental data collected to the edge computing unit, the edge computing unit uses the moving average filtering algorithm to remove noise, and then uses the maximum-minimum normalization method to unify the data with different dimensions into the [0,1] interval to complete data preparation.

[0034] S2. Input the historical environmental data of the box end and the environmental parameter thresholds into a pre-trained decision tree model for classification judgment, and mark the normal environmental data and abnormal environmental data after classification.

[0035] The edge computing unit preprocesses the historical environmental data of the box end and the environmental parameter thresholds and then inputs them into the decision tree model. The data features of this model are various environmental parameters, and the conditional branches are constructed based on the preset environmental parameter thresholds formulated according to a large amount of historical environmental data and equipment operation requirements.

[0036] Specifically, starting from the root node, classification and judgment are performed along the conditional branches according to the values of the environmental parameters in the input data. The decision tree model can classify and make decisions on the data based on the data characteristics and conditional branches. The data characteristics are derived from various environmental parameters collected by the multi-modal environmental perception module. These data are denoised by the moving average filtering algorithm and processed by the maximum-minimum normalization method before entering the model. The conditional branches are constructed based on a large amount of historical data and equipment operation requirements, and data classification and judgment are performed according to different environmental parameter thresholds. The input of the decision tree model is the environmental data presented in the form of a standardized vector after preprocessing, and the output is a structured environmental condition judgment result, including condition classification, abnormal parameter identification, and preliminary risk warning. Subsequently, this preliminary classification result is further compared with the preset normal range thresholds of each environmental parameter. If an environmental parameter exceeds the normal range, it is marked as abnormal, and the environmental condition judgment is refined by comprehensively considering all abnormal parameter situations. Logically, the decision tree model first makes a preliminary classification decision on the environmental data and transmits the abnormal information to the risk index calculation module.

[0037] In this embodiment, the threshold judgment in S1 is actually a pre-rapid judgment link in the data analysis process in S2. Performing the threshold judgment after data collection can detect obvious environmental abnormalities and trigger alarms in the first place. The in-depth analysis in S2 processes and mines the data at a more detailed and comprehensive level. The two form a logical process from rapid preliminary judgment to in-depth analysis, jointly serving the environmental monitoring and risk assessment of the terminal box.

[0038] S3. Train the pre-constructed risk index calculation model by inputting the classified normal environmental data, abnormal environmental data, and their corresponding labels based on a preset loss function to obtain a trained risk index calculation model.

[0039] The fusion decision-making module first denoises and normalizes the data collected by the multi-modal environmental perception module, and then deeply analyzes it using algorithms such as data mining to explore the relationships between environmental parameters. Then, a risk index calculation model is constructed, which is trained and optimized to quantify environmental risks and provide a basis for adjustment decisions. S3 specifically includes: S3.1. Preprocess the classified multiple environmental data.

[0040] Specifically, convert the classified multiple environmental data into a data matrix, perform standardization processing on the data matrix to eliminate the influence of dimensions, and calculate the standardized environmental parameter sample values through the following formula: , where represents the i-th sample value under the j-th standardized environmental parameter, represents the i-th sample value under the j-th environmental parameter to be standardized, represents the j-th environmental parameter.

[0041] S3.2. Based on the linear regression algorithm, construct a risk index calculation model. When constructing the risk index calculation model, initialize the weight of each parameter environment according to the probability of each environmental data under each environmental parameter.

[0042] In S3.2, initializing the weight of each parameter environment according to the probability of each environmental data under each environmental parameter includes: S3.2.1. Calculate the probability of each sample (i.e., each environmental data) under each environmental parameter.

[0043] Specifically, calculate the probability of each sample under each environmental parameter according to the following formula: , where, represents the probability value of the i-th environmental data sample under the j-th environmental parameter, represents the value of the i-th environmental data sample under the j-th environmental parameter, represents the number of environmental data samples under the j-th environmental parameter.

[0044] S3.2.2. Calculate the information disorder degree value of each environmental parameter according to the probabilities of each sample.

[0045] Specifically, calculate the information disorder degree value of each environmental parameter according to the following formula: , where, represents the information disorder degree value of the j-th environmental parameter, represents the probability value of the i-th environmental data sample under the j-th environmental parameter, represents the number of environmental data samples under the j-th environmental parameter.

[0046] S3.2.3. Calculate the information utility value of each environmental parameter based on the information disorder degree value of each environmental parameter.

[0047] Specifically, calculate the information utility value of each environmental parameter according to the following formula: , where, represents the information utility value of the j-th environmental parameter, represents the information disorder degree value of the j-th environmental parameter, The larger it is, the more important the environmental parameter is in the evaluation.

[0048] S3.2.4. Initialize the weights of each environmental parameter in the risk index calculation model according to the information utility value of each environmental parameter: , where, represents the weight of the j-th environmental parameter, represents the information utility value of the j-th environmental parameter, represents the number of environmental parameters.

[0049] S3.3. Based on machine learning, use the historical environmental data and historical device operation status data at the box end to train the risk index calculation model. During the training process, use the backpropagation algorithm to adjust the initial weights of each environmental parameter to obtain a trained risk index calculation model.

[0050] Calculate the preset loss function according to the following formula in 3: , where, , where, represents the loss value, represents the predicted value of the risk index, represents the true value of the risk index, represents the regularization coefficient, represents the number of environmental data samples, represents the number of environmental parameters, represents the weight of the j-th environmental parameter in the risk index calculation model, represents the weight of the i-th environmental data sample.

[0051] In this embodiment, through the preset loss function, the risk index calculation model will pay more attention to high-risk samples during training and maintain the generalization ability through the regularization coefficient, so that the risk index calculation model can more accurately evaluate the true operating environment of the terminal box from the environmental data under different working conditions.

[0052] Specifically, first, collect the environmental data of the terminal box and the corresponding device operation status data under different working conditions as the training set; the working conditions cover various states that may occur during the operation of the terminal box; the environmental data includes the real-time measurement values of n environmental parameters, and the device operation status data is used to determine whether the device is operating normally and the severity of the fault in the current environment, which is used as the actual basis for judging the environmental risk.

[0053] Then, use the backpropagation algorithm in machine learning to optimize the initial weights of the risk index calculation model; input the environmental parameters in the training set into the model, and the model calculates according to the current initial weights , calculate the predicted value of the risk index according to the formula; compare the predicted value with the actual risk situation, calculate the error between the two, and the actual risk situation is determined based on the operating state of the equipment; the backpropagation algorithm propagates the error from the output layer back to the input layer, and continuously adjusts the weights by using the gradient descent method, so that the error gradually decreases until the model reaches the expected prediction performance. Finally, the final environmental weight is obtained.

[0054] In this embodiment, the pre-trained risk index calculation model can discover potential laws that are difficult to directly observe from a large amount of environmental and equipment operating state data under different working conditions, making the prediction more in line with the actual situation, being able to comprehensively and accurately evaluate the true operating environment of the terminal box, providing sufficient data support for subsequent control decisions, and thus realizing more precise and effective environmental regulation.

[0055] S4. Input the multiple preprocessed environmental data in S1 into the trained risk index calculation model in S3 for risk assessment to obtain the risk index value.

[0056] S5. Judge whether the risk index value exceeds the risk threshold. If it exceeds, determine the control mode according to the environmental parameter threshold and the risk index value, and adjust the operating parameters of the equipment in the terminal box according to the determined control mode.

[0057] The fusion decision-making module judges whether the risk index value exceeds the risk threshold (for example, 0.5). If it exceeds, send an adjustment command to the adaptive dynamic adjustment module; the adaptive dynamic adjustment module includes a control module switching unit and an equipment parameter adjustment unit. The control module switching unit is used to determine the control mode according to the environmental parameter threshold and the risk index value after receiving the adjustment command sent by the fusion decision-making module. The equipment parameter adjustment unit is used to adjust the operating parameters of the equipment in the terminal box according to the determined control mode, so as to control factors such as temperature, humidity, gas, light, and vibration in the terminal box, and ensure that the equipment is always in a stable operating environment.

[0058] It can be understood that those skilled in the art set the specific value of the risk threshold according to the actual situation, and this embodiment does not make any restrictions.

[0059] Determining the control mode according to the environmental parameter threshold and the risk index value in S5 includes: S5.1. Determine the control mode switching condition according to each environmental parameter threshold and the risk index value.

[0060] Each environmental parameter has a corresponding environmental parameter threshold. Determining the switching condition according to the environmental parameter threshold and the risk index value is a rule. Here, the risk index value can be a specific value or a value range, which is set according to the actual application situation.

[0061] In this embodiment, the control mode is used to control the device to adjust the environment inside the terminal box. For example, when the gas sensor detects that the concentration of harmful gases exceeds the standard, the system starts the enhanced ventilation equipment, such as turning on the exhaust fan to increase the air exchange volume, so as to reduce the concentration of harmful gases and protect the device from damage; when the temperature and humidity sensor detects that the temperature is too high and the humidity exceeds the normal range, it will control the fan speed to increase, strengthen ventilation and heat dissipation, and at the same time start the dehumidifier to increase the dehumidification power, such as increasing the fan speed from 1000 revolutions per minute to 1500 revolutions per minute and increasing the dehumidifier power from 300W to 500W to adjust the temperature and humidity environment; if the light sensor detects abnormal light, it may control the shading equipment, such as closing the electric shading curtain, to prevent the external light from affecting the internal equipment; when the vibration sensor detects abnormal vibration, it may control the shock absorption device to start, or pause the operation of some vibration-sensitive equipment to avoid damage to the equipment due to vibration.

[0062] S5.2. Use the fuzzy logic algorithm to take the risk index value and multiple environmental data as inputs, and combine the control mode switching conditions for fuzzy reasoning to obtain a comprehensive evaluation result of the environmental state.

[0063] Specifically, convert the input risk index value and multiple environmental data into fuzzy linguistic variables, construct fuzzy sets, convert the input into the membership degrees of the fuzzy sets, and the membership functions can be triangular functions, trapezoidal functions, and Gaussian functions. According to the control mode switching conditions formulated in S5.1, establish expressions in the form of "IF-THEN", and then based on these fuzzy rules, perform logical reasoning on the already fuzzified input variables. For each fuzzy rule, calculate the membership degree of the output fuzzy set according to the membership degree of the input fuzzy set, and then convert the fuzzy set into a more specific fuzzy evaluation result; since the result obtained from fuzzy reasoning is fuzzy, it is necessary to convert the fuzzy evaluation result into a clear and definite numerical value or state description through the centroid method or the maximum membership degree method to obtain a comprehensive evaluation result of the environmental state; the fuzzy rules are expressed in the form of "IF-THEN", describing the relationship between the input risk index value and environmental data and the comprehensive evaluation result of the output environmental state.

[0064] For example, when the input temperature is 45°C, after fuzzification, fuzzy reasoning, and defuzzification, the power of the heater is obtained as 50% (for example, the heater power range is 0 - 100%). The comprehensive evaluation results include the wind speed of the exhaust fan, the power of the heater, the power of the dehumidifier, etc.

[0065] S5.3. Determine the control mode to be switched according to the comprehensive evaluation result.

[0066] Specifically, based on the comprehensive evaluation result of the environmental status, referring to the pre-set corresponding relationship, determine the control mode that needs to be switched in the current situation, and then control the relevant equipment in the terminal box to adjust the environment inside the box.

[0067] For example, the comprehensive evaluation result includes that the heater power is 50%. The control mode switching unit finds the corresponding control mode as 1 according to the comprehensive evaluation result. The device parameter adjustment unit then sets the heater power to 50% according to control mode 1. After the heater receives the control instruction, it adjusts the power to 50%, thereby adjusting the environmental temperature inside the terminal box.

[0068] The method further includes: S6. Cache the pre-processed data in S1 and the analysis result in S4, and upload the cached data to the cloud server when the network between the terminal box and the cloud server is normal; receive remote control instructions and adjust the operating parameters of the equipment in the terminal box according to the remote control instructions.

[0069] The edge computing unit of the cloud-edge collaboration module processes the perception data in real time, judges the environmental conditions, issues instructions and caches the results when there are risks, and uploads the data when the network is normal; the cloud server receives the data for analysis, uses the clustering analysis algorithm to classify the historical data, finds the operation mode rules of the equipment, then builds a management platform to realize remote status monitoring, parameter setting and control instruction issuing operations, and issues remote management instructions to the edge computing unit to realize the collaborative work between the cloud and the edge.

[0070] The method further includes: S7. Dynamically adjust the power supply voltage and current of each module in the terminal box controller based on the reinforcement learning algorithm according to the obtained power supply data, environmental data and equipment load of the terminal box controller.

[0071] The low-power and reliability module dynamically adjusts the module power supply using an intelligent algorithm based on the system power supply and environmental parameters to achieve low-power operation; adopts the methods of hardware redundancy and data verification and retransmission to ensure the uninterrupted operation of the system and the accurate data interaction in case of hardware failures.

[0072] In this embodiment, the deep Q-network (DQN) algorithm is selected as the core model, which includes an evaluation network and a target network, both of which adopt a 3-layer fully connected layer structure, and the number of neurons in each layer is 64, 32, and 16 respectively, and the activation function is the ReLU function. The pre-processed system power supply data, terminal box environmental data and real-time equipment working load data are composed into a multi-dimensional state vector to construct the state space. The action space is defined as a series of different power supply strategies, covering the adjustment of the power supply voltage level and current magnitude of each module.

[0073] Calculate the comprehensive reward function according to the following formula: , wherein, is the energy consumption saving ratio, which is obtained by dividing the difference in energy consumption between the current and the previous moment by the energy consumption at the previous moment; is used to evaluate the stability of the device operation, and is quantified according to the change of the key performance indicators of the device. It is measured by the time interval from detecting the change in the device workload to adjusting the power supply strategy and making the device operate stably; are all weight coefficients, which are used to measure the relative importance of different reward factors in the comprehensive reward.

[0074] Quantifying according to the change of the key performance indicators of the device can determine the key performance indicators related to the operation stability of the equipment in the terminal box. According to the equipment specifications and historical operation data, a normal fluctuation range is set for each indicator. The numerical values of the indicators are monitored in real time. When fluctuating within the normal range, a higher stability reward value is given according to the fluctuation amplitude; if it exceeds the normal range, the reward value is reduced according to the exceeding degree; a higher stability reward value is given when the equipment performance indicators fluctuate within the normal range, otherwise the reward value is reduced.

[0075] For the decision-making adjustment and learning process, the system selects and executes the power supply strategy from the action space according to the ε-greedy strategy under different operating conditions; after execution, the reward value is calculated according to the reward function, and the state, action, reward and next state information are recorded.

[0076] When the number of samples in the experience replay buffer reaches the preset threshold, a batch of samples is randomly selected for training. The evaluation network calculates the Q value of the current state action, the target network calculates the maximum Q value of the next state, and the parameters of the evaluation network are updated by the stochastic gradient descent algorithm by minimizing the loss function. The parameters of the evaluation network are copied to the target network every 100 steps.

[0077] Optionally, the target Q value is calculated according to the following formula: , wherein, represents the target Q value, which is a reference value used to guide the update of the evaluation network parameters in the reinforcement learning process. is the immediate reward value obtained after the system executes the power supply strategy, that is, the reward immediately obtained after executing the action. is the discount factor, and its value range is between 0 and 1. represents the next state transferred to after executing the action. represents all possible actions in the next state. is the Q value calculated by the target network for executing the action in the next state . It represents the maximum Q value among all possible actions in the next state, that is, the optimal action value that can be obtained in the next state.

[0078] Optionally, the loss function is expressed as follows: , where represents the value of the mean squared error loss function, which is used to measure the Q value calculated by the evaluation network and the target Q value The degree of difference between them, represents the number of samples used for training each time, is the Q value calculated by the evaluation network for executing the action in the current state .

[0079] If a high reward value is obtained for executing an action, the system strengthens this decision; if the reward value is low, the strategy is adjusted. Through continuous attempts, obtaining rewards, replaying experiences, and training the model, the system masters the optimal power supply strategy, realizes dynamic adjustment of the power supply voltage and current, and achieves the goals of energy conservation and ensuring the normal operation of equipment.

[0080] Combined with Figure 2 As shown, Embodiment 2 of the present invention provides an adjustment system applied to a terminal box environment controller, which runs the adjustment method applied to the terminal box environment controller provided in Embodiment 1. The system includes a multi-modal environment perception module, a cloud-edge collaboration module, a fusion decision module, and an adaptive dynamic adjustment module. Among them, The multi-modal environment perception module is used to collect various environmental data inside the terminal box.

[0081] The multi-modal environment perception module includes a temperature and humidity sensor unit, a gas sensor unit, a light sensor unit, and a vibration sensor unit; The temperature and humidity sensor unit is used to measure the temperature and humidity inside the terminal box and transmit the measured temperature and humidity data to other modules, enabling the system to monitor in real time whether the temperature and humidity environment inside the terminal box exceeds the threshold range, facilitating the system to take adjustment measures; The gas sensor unit is equipped with several sensors for different harmful gases. The gas sensor unit is responsible for monitoring the concentration changes of harmful gases inside the terminal box and transmitting the monitoring data to the system. When the concentration of harmful gases is detected to exceed the safety threshold, the system will take measures to strengthen ventilation and issue an alarm to avoid the risk of damage to the equipment inside the terminal box; The light sensor unit uses photosensitive elements to monitor the light intensity inside the terminal box in real time. The inside of the terminal box is in a relatively enclosed environment under standard conditions, and the light intensity is in a stable state. However, if the light intensity exceeds the threshold range, it means that external light has entered the terminal box. The light sensor detects the abnormal light situation, transmits it to the system, and issues an alarm to prompt the maintenance personnel to conduct an inspection. The vibration sensor unit continuously senses the vibration condition of the terminal box through a piezoelectric sensor, and obtains the vibration amplitude and frequency parameters inside the terminal box. Under normal operating conditions, the terminal box is in a stable state. When the vibration parameters exceed the threshold range, the system will determine that the current vibration state will affect the equipment based on the collected vibration parameters, and then issue an alarm to prompt the maintenance personnel to conduct maintenance.

[0082] The cloud-edge collaboration module includes an edge computing unit, which is used to preprocess various environmental data collected by the multi-modal environmental perception module. When any environmental data exceeds the corresponding environmental parameter threshold, an alarm is triggered, and the historical environmental data and environmental parameter threshold of the box end are input into a pre-trained decision tree model for classification judgment, and the normal environmental data and abnormal environmental data after classification are marked.

[0083] Optionally, the cloud-edge collaboration module further includes a cloud server unit, which is used to receive and store the preprocessed data uploaded by the edge computing unit, classify the preprocessed data using a clustering analysis algorithm to obtain the equipment operation mode rules, and display the equipment operation mode rules to the user. At the same time, it receives the remote control instructions from the user for the terminal box environment controller. The remote control instructions include the setting of terminal box environmental parameters, and the operation parameters of the equipment inside the terminal box are adjusted according to the remote control instructions.

[0084] The cloud server unit is responsible for receiving the data uploaded by the edge computing unit, classifying the historical data using a clustering analysis algorithm to find the equipment operation mode rules under different environmental parameter combinations, and building a Web-based management platform to provide RESTful API interfaces for interaction with the user, realizing remote status monitoring, parameter setting, and control instruction issuing operations for the terminal box. In this embodiment, when using the clustering analysis algorithm to process historical data, the Density Peaks Clustering (DPC) algorithm is adopted. Compared with traditional clustering algorithms, this algorithm does not require pre-setting the number of clusters, and is more suitable for dealing with the complex and changeable parameter combinations in the terminal box environment data. The data uploaded by the edge computing unit includes environmental parameters such as temperature, humidity, harmful gas concentration, light intensity, vibration, etc. collected by the multi-modal environment perception module and the corresponding device operation status data. First, these data are pre-processed. Outliers and duplicate data are removed through data cleaning, and then feature extraction is performed to extract the change trends of environmental parameters and the fluctuation characteristics of device operation status in different time periods. Then, the processed data is input into the DPC algorithm. Based on the local density and relative distance between data points, the algorithm automatically identifies the cluster centers in the data, thereby classifying the historical data according to different environmental parameter combinations.

[0085] In this embodiment, when building a Web-based management platform, a front-end and back-end separation architecture is adopted. The front end uses the Vue.js framework to build the user interface, where users can view the values and change curves of various environmental parameters in the terminal box in real time, and intuitively understand the operation status of the device. The back end uses the Spring Boot framework to process the requests sent by the front end and interact with the database. When providing RESTful API interfaces, the REST architecture style is strictly followed, and the interface specifications are defined. For the operation of obtaining the real-time environmental parameters of the terminal box, a GET request interface is defined. The front end requests data from the server through this interface, and the server queries the latest environmental parameters from the database according to the request and returns them to the front end. For the operation of setting the control parameters of the terminal box, a PUT request interface is defined. The front end sends the parameters set by the user to the server through this interface. After the server validates and processes the parameters, it issues the remote management instruction to the edge computing unit through the communication unit to achieve remote control of the terminal box.

[0086] The fusion decision-making module is used to input the classified normal environmental data, abnormal environmental data and their corresponding labels sent by the edge computing unit into a pre-constructed risk index calculation model for training based on a preset loss function to obtain a trained risk index calculation model; and input the multiple environmental data pre-processed by the edge computing unit into the trained risk index calculation model for risk assessment to obtain a risk index value; determine whether the risk index value exceeds the risk threshold, and in the case of exceeding, send an adjustment command to the adaptive dynamic adjustment module.

[0087] The fusion decision-making module includes a data pre-processing unit, a data analysis unit and a risk index calculation unit; The data preprocessing unit is responsible for denoising and normalizing the collected data to improve data quality and provide a basis for subsequent data analysis. The data analysis unit uses data mining and machine learning algorithms to deeply analyze the preprocessed data, and then uses association rule algorithms to mine the potential relationships and laws between environmental parameters to provide a basis for risk assessment.

[0088] The adaptive dynamic adjustment module includes a control module switching unit and a device parameter adjustment unit. The control module switching unit is used to determine the control mode according to the environmental parameter threshold and the risk index value after receiving the adjustment command sent by the fusion decision module. The device parameter adjustment unit is used to adjust the operating parameters of the devices in the terminal box according to the determined control mode.

[0089] Optionally, the adjustment system further includes a low-power and reliability module, which includes an adaptive power regulation unit, a hardware device redundancy switching unit, and a fault intelligent fault tolerance processing unit. Among them, The adaptive power regulation unit is used to dynamically adjust the power supply voltage and current of each module in the terminal box controller based on the reinforcement learning algorithm according to the obtained power data, environmental data, and device load of the terminal box controller.

[0090] In this embodiment, when the device is in a low-load state at night and the environmental parameters are stable, other modules except the environmental perception module are controlled to enter the deep sleep mode, reducing the system power consumption by 50%. When the device workload increases during the day, the relevant modules are awakened and the normal power supply is restored.

[0091] The terminal box environmental controller includes a main hardware and a backup hardware. The hardware device redundancy switching unit is used to monitor the synchronization status and data backup of the main hardware and the backup hardware in real time, and automatically switch to the backup hardware when the main hardware fails to ensure that the system's processing and regulation of the terminal box environmental data are uninterrupted.

[0092] The fault intelligent fault tolerance processing unit is used to use the cyclic redundancy check algorithm to check the data collected by the multi-modal environmental perception module and transmitted to the edge computing unit during the data transmission process. When a transmission error is detected in the data, a retransmission instruction is generated, and the data sending end (such as the multi-modal environmental perception module) is requested to retransmit the data through the cloud server unit. In this way, the accuracy of data interaction between system modules is ensured.

[0093] In this embodiment, the main hardware refers to a set of hardware devices that undertake core data processing, calculations, and control instruction execution tasks when the system is operating normally; the backup hardware refers to hardware devices with the same or partially equivalent functions as the main hardware, which can promptly take over the main hardware work when the main hardware fails to maintain the normal operation of the system. The purpose of the fast fault detection algorithm is to quickly detect and trigger the switching of backup hardware at the moment of failure of the main hardware, ensuring that the system processes and regulates the terminal box environment data uninterruptedly; the algorithm is based on data mining technology and realizes fault detection by real-time monitoring and analysis of key parameters during the operation of the main hardware.

[0094] Regarding the system in the above embodiment, the specific manner in which each unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0095] Embodiment 3 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is loaded into the processor, the adjustment method for the terminal box environment controller described in Embodiment 1 is implemented.

[0096] Embodiment 4 of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the adjustment method applied to the terminal box environment controller according to embodiment 1 is implemented.

[0097] Compared with the prior art, the beneficial effects of the present invention include at least: 1. Comprehensive environmental perception: The multimodal environmental perception module of the present invention integrates multiple sensor units such as temperature and humidity, gas, light, and vibration. Compared with the traditional single temperature and humidity monitoring, it can collect multi-dimensional environmental data in the terminal box in an all-round and high-precision manner. This enables the system to have a more comprehensive and accurate understanding of the real environmental conditions of the terminal box, providing a rich and reliable data basis for subsequent decision-making and adjustment, and effectively avoiding equipment operation risks caused by the lack of environmental information.

[0098] 2. Intelligent decision-making: The fusion decision-making module abandons the traditional simple threshold judgment method through data preprocessing, in-depth data analysis and scientific risk index calculation. It can not only explore the potential relationship and rules between different environmental parameters, but also accurately calculate the risk index according to the impact of each parameter on the environment. It provides a scientific, quantitative and comprehensive basis for regulatory decisions, greatly improves the accuracy and foresight of decision-making, and enables the system to predict and respond to potential environmental risks in advance.

[0099] 3. Adaptive dynamic regulation: The adaptive dynamic regulation module can intelligently switch between multiple control modes based on the risk index and real-time environmental parameters, and precisely and dynamically adjust the operating parameters of the regulating device. Compared with the traditional regulation method with fixed modes and parameters, the present invention can better adapt to the complex and changeable environmental requirements of the terminal box, achieve more precise and efficient environmental regulation, and ensure that a stable environment suitable for the operation of the equipment is always maintained inside the terminal box.

[0100] 4. Management mode coordination: The cloud-edge coordination module realizes the efficient coordination of local real-time decision-making and remote management. The edge computing unit ensures the timeliness of local data processing and real-time decision-making, and the system can still operate normally even in the case of network anomalies; the cloud server unit provides powerful remote data storage, analysis, and management functions, facilitating users to monitor and manage the terminal box anytime and anywhere. This coordination mode significantly improves the management efficiency and convenience of the system, breaking the limitation of the separation of local and remote functions in the traditional system.

[0101] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution is prior or posterior, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0102] The present disclosure may be a system, method, and / or computer program product. The computer program product may include a computer-readable storage medium having thereon computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0103] The computer-readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device. The computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device, such as a punched card or raised structures in grooves storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage medium used herein is not construed as an instantaneous signal itself, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through an optical fiber cable), or electrical signals transmitted through wires.

[0104] The computer-readable program instructions described herein can be downloaded to various computing / processing devices from a computer-readable storage medium or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0105] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state-setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the present disclosure.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent substitutions can still be made to the specific embodiments of the present invention, and any modification or equivalent substitution that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A method for adjusting a terminal box environment controller, characterized in that: include: Collect various environmental data in the terminal box and pre-process the collected environmental data. When any environmental data exceeds the corresponding environmental parameter threshold, an alarm is triggered; Input the historical environmental data and environmental parameter thresholds of the box end into the pre-trained decision tree model for classification and judgment, and mark the classified normal environmental data and abnormal environmental data; Based on a preset loss function, the classified normal environment data and abnormal environment data and their corresponding labels are input into a pre-built risk index calculation model for training to obtain a trained risk index calculation model; The risk index calculation model trained by inputting pre-processed multiple environmental data is used to conduct risk assessment and obtain the risk index value; Determine whether the risk index value exceeds the risk threshold. If it exceeds the risk threshold, determine the control mode according to the environmental parameter threshold and the risk index value, and adjust the operating parameters of the equipment in the terminal box according to the determined control mode.

2. The adjustment method for a terminal box environment controller according to claim 1, characterized in that: A risk index calculation model is constructed based on a linear regression algorithm. When constructing the risk index calculation model, the weight of each parameter environment is initialized according to the probability of each environmental data under each environmental parameter.

3. The adjustment method for a terminal box environment controller according to claim 2, characterized in that: The weight of each parameter environment is initialized according to the probability of each environmental data under each environmental parameter, including: Calculate the probability of each environmental data under each environmental parameter; Calculate the information disorder degree value of each environmental parameter according to the probability of each environmental data; The information utility value of each environmental parameter is calculated based on the information disorder degree value of each environmental parameter; The weights of each environmental parameter in the risk index calculation model are initialized according to the information utility value of each environmental parameter.

4. The adjustment method for a terminal box environment controller according to claim 3, characterized in that: The weights of each environmental parameter in the risk index calculation model are initialized according to the information utility value of each environmental parameter as follows: ， in, represents the weight of the jth environmental parameter, represents the information utility value of the jth environmental parameter, Indicates the number of environment parameters.

5. The adjustment method for a terminal box environment controller according to claim 2, characterized in that: The preset loss function is calculated as follows: ， in, represents the loss value, represents the predicted value of risk index, represents the true value of the risk index, represents the regularization coefficient, represents the number of environmental data samples, represents the number of environmental parameters, represents the weight of the jth environmental parameter in the risk index calculation model, Represents the weight of the i-th environmental data sample.

6. The adjustment method for a terminal box environment controller according to any one of claims 1 to 5, characterized in that: The control mode is determined based on the environmental parameter threshold and risk index value, including: Determine the control mode switching condition according to each environmental parameter threshold and risk index value; Using fuzzy logic algorithm, the risk index value and multiple environmental data are used as input, and fuzzy reasoning is performed in combination with the control mode switching conditions to obtain a comprehensive evaluation result of the environmental status; The control mode to be switched is determined based on the comprehensive evaluation results.

7. The adjustment method for a terminal box environment controller according to claim 6, characterized in that: The method further comprises: Receive remote control instructions and adjust the operating parameters of the equipment in the terminal box according to the remote control instructions.

8. The adjustment method for a terminal box environment controller according to claim 6, characterized in that: The method further comprises: Based on the reinforcement learning algorithm, the power supply voltage and current of each module in the terminal box controller are dynamically adjusted according to the acquired terminal box controller power data, environmental data and equipment load.

9. A regulating system applied to a terminal box environment controller using the regulating method applied to a terminal box environment controller according to any one of claims 1 to 8, characterized in that: The system comprises: Multimodal environmental perception module, used to collect various environmental data in the terminal box; The cloud-edge collaboration module includes an edge computing unit, which is used to pre-process the various environmental data collected by the multimodal environmental perception module, trigger an alarm when any environmental data exceeds the corresponding environmental parameter threshold, input the box-side historical environmental data and environmental parameter threshold into the pre-trained decision tree model for classification and judgment, and mark the classified normal environmental data and abnormal environmental data; A fusion decision module is used to input the classified normal environment data and abnormal environment data and their corresponding labels sent by the edge computing unit into a pre-built risk index calculation model for training based on a preset loss function to obtain a trained risk index calculation model; and input a variety of environmental data pre-processed by the edge computing unit into the trained risk index calculation model for risk assessment to obtain a risk index value; determine whether the risk index value exceeds the risk threshold, and if so, send an adjustment command to the adaptive dynamic adjustment module; The adaptive dynamic adjustment module includes a control module switching unit and an equipment parameter adjustment unit. The control module switching unit is used to receive the adjustment command sent by the fusion decision module, and then determine the control mode according to the environmental parameter threshold and the risk index value. The equipment parameter adjustment unit is used to adjust the operating parameters of the equipment in the terminal box according to the determined control mode.

10. The regulating system for a terminal box environment controller according to claim 9, characterized in that: The cloud-edge collaboration module also includes a cloud server unit, which is used to receive and store the preprocessed data uploaded by the edge computing unit, classify the preprocessed data using a clustering analysis algorithm, obtain the equipment operation mode rules, and display the equipment operation mode rules to the user, and at the same time receive the user's remote control instructions for the terminal box environment controller, the remote control instructions include terminal box environment parameter settings, and adjust the operating parameters of the equipment in the terminal box according to the remote control instructions.

11. The regulating system for a terminal box environment controller according to claim 9, characterized in that: The regulation system also includes a low power consumption and reliability module, which includes an adaptive power supply control unit, a hardware device redundancy switching unit and a fault intelligent fault tolerance processing unit; wherein, The adaptive power control unit is used to dynamically adjust the power supply voltage and current of each module in the terminal box controller based on the acquired terminal box controller power data, environmental data and equipment load based on the reinforcement learning algorithm; The terminal box environment controller includes main hardware and backup hardware. The hardware device redundancy switching unit is used to perform synchronous status monitoring and data backup on the main hardware and backup hardware in real time, and automatically switch to the backup hardware when the main hardware fails. The fault intelligent fault-tolerant processing unit is used to verify the data collected by the multimodal environment perception module and transmitted to the edge computing unit using a cyclic redundancy check algorithm during the data transmission process, and generate a retransmission instruction when a data transmission error is detected, and request the multimodal environment perception module to resend the data through the cloud server unit.

12. An electronic device, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-8.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images