Intelligent monitoring system based on heat flow multi-dimensional characteristics and cross-scene application method
Patent Information
- Application Number
- CN202511245620.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-02
- Publication Date
- 2025-12-19
AI Technical Summary
Existing heat flow monitoring technologies suffer from issues such as signal distortion, limited monitoring dimensions, and poor cross-scenario applicability, resulting in response delays, high false alarm rates, and high costs associated with repetitive development.
By employing a distributed heat flux sensor array, an intelligent heat flux transmitter, and a dynamic risk assessment module, a cross-scenario intelligent monitoring system is constructed through multi-dimensional feature fusion and adaptive control. This system includes signal conditioning, feature normalization, and risk index calculation, enabling reliable signal transmission and cross-scenario adaptive decision-making.
It enables reliable transmission of weak heat flow signals with high impedance, deeply mines multi-dimensional features, improves the accuracy of monitoring and cross-scenario adaptability, and reduces false alarm rate and maintenance cost.
Smart Images
Figure CN121170949A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of intelligent sensing and monitoring, and particularly relates to a monitoring system based on intelligent analysis, evaluation and regulation of time domain, space domain and steady-state multi-dimensional characteristics of heat flow signals and a cross-scene application method. BACKGROUND
[0002] In key fields such as industrial production, power systems, new energy and building fire protection, accurate monitoring of the thermal state of equipment or environment is a core link to ensure safety and improve efficiency. Currently, thermal quantity detection mainly relies on temperature sensors such as thermistors and thermocouples. However, such temperature measurement-based technologies have inherent limitations. First, response delay: temperature is a macroscopic manifestation of heat accumulation, and temperature sensors must wait for a significant temperature rise at the monitoring point before responding, which can be fatal in scenarios requiring fast response, such as fire warning, and can miss the valuable initial intervention window. Second, single monitoring dimension: single temperature threshold judgment logic is easily disturbed by environmental temperature fluctuations or short-term normal overload of equipment, resulting in high false alarm rate and inability to accurately reflect the dynamic development process and spatial distribution of thermal events.
[0003] Heat flow, as a physical quantity representing the amount of heat per unit area per unit time, can more directly and quickly reflect the heat transfer rate and dynamic changes. Heat flow sensor response speed can reach milliseconds, providing a physical basis for predictive maintenance and early warning. For example, in battery thermal runaway monitoring, heat flow signals can capture abnormal signs several minutes earlier than temperature signals. Despite the significant advantages of heat flow monitoring, it still faces three major challenges in engineering applications:
[0004] (1) Signal chain vulnerability: heat flow sensors output weak voltage signals of microvolts to millivolts, with internal resistance up to kilo-ohms (kΩ) to mega-ohms (MΩ). This high impedance, weak signal is easily affected by electromagnetic interference, system input bias current and long cable voltage division effect during transmission, resulting in severe signal distortion. The lack of standardized intelligent transducers designed specifically for such signal characteristics on the market makes it difficult to ensure signal fidelity and reliability.
[0005] (2) Dynamic parameter value not fully exploited: traditional heat flow measurement applications mostly focus on measuring steady-state heat flux, such as thermal resistance testing of building materials or fireproof clothing. The time domain dynamic characteristics represented by heat flow rate dQ / dt and the spatial distribution characteristics represented by heat flow gradient ΔQ have not been systematically developed and utilized in abnormal event diagnosis and fault location.
[0006] (3) Lack of cross-scene adaptive capability: Different application scenarios have different focuses on thermal risks. For example, battery thermal runaway warning focuses on the sharp change of heat flow (dQ / dt), industrial equipment local fault monitoring focuses on abnormal hot spots ΔQ, and long-term load monitoring of power cables focuses on the sustained over-limit of heat flow absolute value |Q|. Existing monitoring solutions are usually developed for a single scene, and lack a general architecture that can adaptively adjust strategies and achieve knowledge transfer, resulting in repeated development and high maintenance costs.
[0007] Therefore, there is an urgent need for a new intelligent monitoring technology that not only solves the reliable transmission problem of weak heat flow signals with high impedance, but also deeply integrates the multi-dimensional features of heat flow and builds an intelligent decision-making framework that can adapt to different application scenarios. SUMMARY
[0008] Therefore, the purpose of the present application is to provide an intelligent monitoring system based on multi-dimensional features of heat flow and a cross-scene application method to solve the problems of signal distortion, single monitoring dimension and poor cross-scene generality in the background art.
[0009] To achieve the above-mentioned purpose, the present application provides the following technical solutions:
[0010] An intelligent monitoring system based on multi-dimensional features of heat flow, comprising:
[0011] A heat flow detection unit composed of a distributed heat flow sensor array for real-time acquisition of heat flow density data at multiple points in the monitoring area;
[0012] An intelligent heat flow transmitter connected to the heat flow detection unit for converting the millivolt-level, high-impedance weak heat flow signals collected by the heat flow sensor array into high-reliability digital or analog signals;
[0013] A dynamic risk assessment module connected to the intelligent heat flow transmitter for receiving the high-reliability digital or analog signals and calculating a comprehensive risk index representing the current monitoring state based on a pre-set risk index model; and
[0014] A cross-scene regulation unit connected to the dynamic risk assessment module for switching threshold strategies for risk assessment according to different application scenarios and triggering graded early warning or device regulation instructions according to the level of the comprehensive risk index.
[0015] Further, the risk index model on which the dynamic risk assessment module calculates the comprehensive risk index is:
[0016]
[0017] wherein, the Risk_index is the comprehensive risk index; is a normalized time-domain dynamic feature, the time-domain dynamic feature is a heat flow rate dQ / dt; |Q| norm is a normalized steady-state feature, the steady-state feature is an absolute value of heat flow |Q|; (ΔQ) norm is a normalized spatial distribution feature, the spatial distribution feature is a heat flow difference ΔQ between different nodes in the monitoring area; α, β, γ are dynamic weight coefficients, and α+β+γ=1.
[0018] Further, the cross-scene regulation unit is further configured to adaptively adjust the dynamic weight coefficients according to application scene characteristics:
[0019] In a transient risk scene, the weight coefficient α corresponding to the time-domain dynamic feature is increased;
[0020] In a device local overheating scene, the weight coefficient γ corresponding to the spatial distribution feature is increased;
[0021] In a long-term stable operation scene, the weight coefficient β corresponding to the steady-state feature is increased.
[0022] Further, the intelligent heat flow transmitter comprises:
[0023] a signal conditioning module, configured to filter, signal follow, and differentially amplify the weak heat flow signal to output an analog signal; and
[0024] an embedded processing unit, configured to convert the analog signal into a digital signal and transmit the digital signal through a communication interface.
[0025] Further, the signal conditioning module comprises:
[0026] a filtering circuit, configured to filter out high-frequency common-mode noise and differential noise in the weak heat flow signal;
[0027] a signal following circuit, connected to an output end of the filtering circuit, configured to improve signal driving capability; and
[0028] a differential amplification circuit, connected to an output end of the signal following circuit, configured to amplify the signal.
[0029] Further, the measurement mode of the heat flow detection unit comprises a contact measurement mode or a radiation measurement mode;
[0030] The contact measurement mode directly contacts a measured object surface through a flexible or rigid substrate sensor;
[0031] The radiation measurement mode keeps a preset distance from the measured object through a sensor, and collects thermal radiation energy emitted by the target object in a non-contact manner.
[0032] An intelligent regulation method based on heat flow multi-dimensional characteristics, comprising the following steps:
[0033] Through a distributed heat flow sensor array, heat flow density data at multiple points in the monitoring area is collected.
[0034] Through an intelligent transmission circuit, the collected weak heat flow signal is converted into a high-reliability digital signal or an analog signal.
[0035] The high-reliability digital signal or analog signal corresponding to the time-domain dynamic characteristics is calculated in real time.
[0036] dQ / dt, spatial distribution characteristics ΔQ, and steady-state characteristics |Q|.
[0037] According to the type of the target application scenario, the dynamic weight coefficients α, β, and γ used to construct the comprehensive risk index model are adaptively adjusted.
[0038] Based on the dynamic weight coefficients and the time-domain dynamic characteristics, spatial distribution characteristics, and steady-state characteristics, a comprehensive risk index is calculated.
[0039] When the comprehensive risk index exceeds a preset threshold, a hierarchical response strategy is triggered.
[0040] Further, before calculating the comprehensive risk index, the time-domain dynamic characteristics, spatial distribution characteristics, and steady-state characteristics are normalized.
[0041] Further, the step of adaptively adjusting the dynamic weight coefficients according to the type of the target application scenario comprises:
[0042] When the application scenario is a transient risk scenario, the weight coefficient α corresponding to the time-domain dynamic characteristics is increased.
[0043] When the application scenario is a device local overheating scenario, the weight coefficient γ corresponding to the spatial distribution characteristics is increased.
[0044] When the application scenario is a long-term stable operation scenario, the weight coefficient β corresponding to the steady-state characteristics is increased.
[0045] Further, the method uses machine learning algorithms to train and optimize the dynamic weight coefficients α, β, and γ using historical monitoring data and regulation effect feedback.
[0046] The beneficial effects of the present application are:
[0047] Additional advantages, objects, and features of the application will be apparent to those skilled in the art upon examination of the following specification. It is intended that the application not be limited by any of the recited embodiments, but that the claims be construed as embracing all the features and advantages within their scope. BRIEF DESCRIPTION OF DRAWINGS
[0048] In order to make the objectives, technical solutions and advantages of the present application clearer, the preferred embodiments of the present application will be described in detail below with reference to the accompanying drawings, in which:
[0049] Figure 1 The schematic diagram of the intelligent monitoring system architecture based on thermal flow multi-dimensional characteristics according to the present application;
[0050] Figure 2 The structural schematic diagram of the integrated radiation type thermal flow detection device according to the present application;
[0051] Figure 3 The structural schematic diagram of the integrated contact type thermal flow detection device according to the present application;
[0052] Figure 4 The schematic diagram of the internal analog signal conditioning circuit of the intelligent thermal flow transmitter according to the present application;
[0053] Figure 5 The schematic diagram of the communication interface circuit according to the present application;
[0054] Figure 6 The schematic diagram of the embedded processing unit core single-chip microcomputer according to the present application;
[0055] Figure 7 The flow chart of the risk assessment and cross-domain regulation decision according to the present application.
[0056] Reference signs: measured object 101, thermal radiation receiving layer 202, thermal flow sensing array layer 203, intelligent transmitting circuit 204, external protective shell 205, thermal flow detection area 301, interface adaptation layer 302, thermal flow sensor body 303, intelligent transmitting circuit 304, external protective shell 305. DETAILED DESCRIPTION
[0057] The present application is further explained in the following detailed description with reference to the accompanying drawings, wherein:
[0058] The drawings are only used for illustrative description, and the representation is a schematic diagram, not a physical diagram, and cannot be understood as a limitation on the present application; in order to better illustrate the embodiments of the present application, some components in the drawings are omitted, enlarged or reduced, and do not represent the size of the actual product; it is understandable for those skilled in the art that some well-known structures and their descriptions in the drawings can be omitted.
[0059] The same or similar reference numerals in the drawings of the embodiments of the present application correspond to the same or similar components; in the description of the present application, it should be understood that if the terms "upper", "lower", "left", "right", "front", "back" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, therefore the terms describing the positional relationship in the drawings are only used for illustrative description, and cannot be understood as a limitation on the present application, for those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.
[0060] Referring to Figure 1 , the intelligent monitoring system described in the present application mainly consists of three levels: heat flow detection unit, intelligent transmitter and intelligent control layer.
[0061] 1. Heat flow detection unit
[0062] This unit is the data source of the system, responsible for directly collecting heat flow signals. According to the different monitoring objects, two measurement modes can be adopted:
[0063] Contact measurement mode: as shown in Figure 3 For scenarios that require accurate measurement of solid surface heat flow, a contact heat flow detection device is used. The core of the device is a heat flow sensor body 303), below which an interface adaptation layer 302) is provided, such as thermal conductive silicone grease or liquid metal, used to fill the micro voids between the sensor and the measured surface 301), to ensure that the contact thermal resistance is less than 0.01K·m 2W, thus ensuring the accuracy of the measurement. The sensor substrate can be flexible to fit curved surfaces (such as pipes, batteries), or rigid for flat surfaces (such as switch cabinet panels).
[0064] Radiation measurement mode: as shown in Figure 2 , for high temperature, high pressure or inconvenient contact monitoring objects, a radiation heat flow detection device is used. This device does not need to be in contact with the measured object 101), and through the directional focusing structure and infrared filter in the heat radiation receiving layer 202), it selectively receives the heat radiation energy of a specific area and waveband, which is then converted into an electrical signal by the heat flow sensing array layer 203). The system can automatically calibrate the field of view angle according to the monitoring distance, ensuring the accuracy of the measurement.
[0065] 2. Intelligent heat flow transmitter
[0066] This unit is the core of solving the problem of signal vulnerability. It can be integrated with the heat flow detection unit (such as 204 in Figure 2 and 304 in Figure 3 ), or it can be a separate design connected by an anti-interference cable. Its internal circuit is shown in Figure 4 , which mainly includes:
[0067] Signal conditioning module:
[0068] Filter circuit: the raw signal output by the sensor first enters the passive low-pass filter circuit composed of resistors R1, R13 and capacitors C10, C11, etc. This circuit can effectively filter out high-frequency differential noise and common-mode noise induced by space electromagnetic fields, protecting the subsequent amplification circuit.
[0069] Signal following circuit: the filtered signal enters the dual-op-amp voltage follower composed of operational amplifiers U1, U3. This circuit has extremely high input impedance and extremely low output impedance, which can match the impedance of high-impedance sensor signals, avoid signal voltage division attenuation, and provide strong driving capability for the subsequent circuit.
[0070] Differential amplification circuit: the signal output by the signal follower is sent to the differential amplification circuit composed of operational amplifier U2 and peripheral resistors. This circuit accurately amplifies the differential signal at the microvolt or millivolt level, while further suppressing common-mode interference.
[0071] Embedded processing unit: as shown in Figure 1 and Figure 6 , the conditioned and amplified analog signal can be directly output as an analog voltage / current signal, or it can be sent to an ADC (analog-to-digital converter) for digitization. The core controller (single-chip microcomputer, such as the STM32 series) processes the digital signal and transmits it through various communication interfaces (such as RS485, 4G or LoRa, etc.Figure 5 The RS485 interface shown) sends data to the intelligent control layer.
[0072] 3. Intelligent control layer
[0073] This layer is the "brain" of the system, responsible for data analysis, decision-making, and control. Its core is the dynamic risk assessment module and the cross-scene regulation unit. Its decision-making process is shown in Figure 7
[0074] Feature calculation and normalization: After receiving data from the transmitter, the module calculates three core features in real time: time-domain dynamic feature dQ / dt, steady-state feature |Q|, and spatial distribution feature ΔQ. To eliminate the dimension effect, the module normalizes these three features:
[0075] Time-domain feature normalization: where K t is the preset critical safety value of rapid heat flow change.
[0076] Steady-state feature normalization: where K q is the maximum heat flow range monitored by the system or the upper limit of the allowed steady-state operation.
[0077] Spatial feature normalization: where K s is the maximum heat gradient difference allowed between adjacent sensors.
[0078] Comprehensive risk index calculation: The module uses the following comprehensive risk index model for evaluation:
[0079]
[0080] where α, β, γ are dynamic weight coefficients, satisfying α+β+γ=1.
[0081] The system can automatically identify scene characteristics and dynamically adjust the weight coefficients.
[0082] Scene matching: By analyzing the macro features of heat flow data (such as change rate, fluctuation range, spatial consistency, etc.), the system matches the current state to the preset scene library, such as "transient mutation feature", "local overheating feature" or "thermal imbalance feature".
[0083] Weight adjustment: After successful matching, the system automatically adjusts the weights. For example, if it is identified as an electrical short circuit precursor (transient mutation), the time-domain weight α is increased to enhance the sensitivity to dQ / dt; if it is identified as bearing abnormal heating (local overheating), the spatial weight γ is increased to strengthen the monitoring of ΔQ; if it is identified as long-term overload (thermal imbalance), the steady-state weight β is increased to focus on the monitoring of |Q|.
[0084] Machine learning optimization: In a preferred embodiment, the system integrates a machine learning module. This module continuously collects historical monitoring data and regulation results (such as whether the alarm is accurate, whether the regulation is effective, etc.) as training samples. Through supervised learning or reinforcement learning algorithms, the system can automatically and continuously optimize the weight coefficients a, b, g and the thresholds of each level of alarm, realize the self-learning and self-evolution of the model, and continuously improve the accuracy and self-adaptability of the monitoring.
[0085] Example 1: Thermal anomaly monitoring of high-voltage power transformation equipment
[0086] This embodiment demonstrates the application of the present application in the field of power safety.
[0087] Deployment and collection: At the bus connection of the high-voltage switch cabinet, the contact of the circuit breaker and other key parts, install flexible and pasted thermal flow sensors (contact measurement mode). The intelligent transmitter will condition, amplify and digitize the signals collected by the sensor, and then transmit them to the monitoring host through the RS485 bus.
[0088] Parameter configuration: In this scenario, loose connection or insulation aging may cause rapid temperature rise, which belongs to the risk of considering both steady state and transient state. Therefore, the weight coefficients are set as follows: a = 0.5 (pay attention to transient changes), b = 0.3 (pay attention to long-term overload), g = 0.2 (pay attention to local poor contact). The normalized parameters are set according to the equipment safety regulations.
[0089] Method execution: The system continuously executes the method: collects data, signal conversion, calculates three normalized thermal flow characteristics, and calculates Risk_index according to the above weight.
[0090] Hierarchical response: When Risk_index exceeds 0.7, the system triggers a first-level warning and prompts the operation and maintenance personnel to pay attention; when it exceeds 0.8, it triggers a second-level warning and automatically starts the forced cooling system; when it exceeds 0.9, it triggers a third-level warning and sends a "suggested power-off" instruction to the main control system to prevent the accident from expanding.
[0091] Example 2: Early warning system for building fire
[0092] This embodiment demonstrates the application of the present application in the field of fire safety.
[0093] Deployment and collection: In the data center or warehouse roof, arrange the radiation type thermal flow sensor in a 3m x 3m grid (radiation measurement mode) to monitor the overall thermal radiation change of the space; at the same time, install flexible and pasted sensors (contact measurement mode) at high-risk points such as cable bridge, power distribution cabinet, etc.
[0094] Parameter configuration: The typical feature of the early stage of fire is the rapid increase of heat flux. Therefore, this scenario is identified as a "transient risk scenario", and the system adaptively adjusts the weights as: a = 0.7, β = 0.2, γ = 0.1, greatly improving the sensitivity to dQ / dt.
[0095] Hierarchical response: When Risk_index > 0.6, the system judges that there is a potential risk, and starts the audible and visual alarm and notifies the monitoring center; when Risk_index > 0.8, it is judged that it is the early stage of fire, and the smoke exhaust system is started; when Risk_index > 0.9, it is judged that there is an open fire, and the automatic sprinkler system is triggered and the emergency evacuation broadcast is started.
[0096] Example 3: Intelligent greenhouse environment regulation
[0097] Deployment and collection: A mixed array of sensors is uniformly arranged in the greenhouse to monitor the heat flux distribution on the surface of the crops and in the space.
[0098] Parameter configuration: The core of greenhouse regulation is to ensure the uniformity of the environment and prevent local overheating or overcooling. Therefore, this scenario is identified as a "thermal imbalance feature" or "device local overheating scenario". The system adaptively adjusts the weights as: γ = 0.6, β = 0.3, a = 0.1, focusing on the spatial heat flux difference ΔQ.
[0099] Regulation strategy: The system performs zoned precise regulation according to the heat flux spatial distribution map. For example, if it is monitored that the ΔQ of region A is significantly higher than that of other regions, the ventilation or shading system of that region is turned on to achieve on-demand regulation and save energy.
[0100] Finally, it should be pointed out that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the purpose and scope of the present application, and all should be included in the scope of the claims of the present application.
Claims
1. An intelligent monitoring system based on multi-dimensional characteristics of heat flow, characterized in that: The application relates to a heat flow detection system, comprising: a heat flow detection unit composed of a distributed heat flow sensor array, which is used for collecting heat flow density data of multiple points in a monitoring area in real time; an intelligent heat flow transmitter connected with the heat flow detection unit, which is used for converting weak heat flow signals of millivolt level and high impedance collected by the heat flow sensor array into digital signals or analog signals with high reliability; a dynamic risk assessment module connected with the intelligent heat flow transmitter, which is used for receiving the digital signals or analog signals with high reliability and calculating a comprehensive risk index representing a current monitoring state based on a preset risk index model in real time; and a cross-scene regulation unit connected with the dynamic risk assessment module, which is used for switching threshold strategies of risk assessment according to different application scenes and triggering graded early warning or equipment regulation instructions according to the grade of the comprehensive risk index. The risk index model for calculating the comprehensive risk index by the dynamic risk assessment module comprises: The cross-scene regulation unit is further used for adaptively adjusting the dynamic weight coefficients according to application scene characteristics: In a transient risk scene, the weight coefficient alpha corresponding to the time-domain dynamic feature is increased; In a device local overheating scene, the weight coefficient gamma corresponding to the spatial distribution feature is increased; In a long-term stable operation scene, the weight coefficient beta corresponding to the steady-state feature is increased. 2.The intelligent monitoring system based on heat flow multi-dimensional features according to claim 1, characterized in that: The intelligent heat flow transmitter comprises: wherein, the Risk_index is the comprehensive risk index; is a normalized time-domain dynamic feature, the time-domain dynamic feature is a heat flow rate of change dQ / dt; |Q| norm is a normalized steady-state feature, the steady-state feature is a heat flow absolute value |Q|; (ΔQ) norm is a normalized spatial distribution feature, the spatial distribution feature is a heat flow difference value ΔQ between different nodes in the monitoring area; α, β, γ are dynamic weight coefficients, and α+β+γ=1. 3.The intelligent monitoring system based on heat flow multi-dimensional features of claim 2, wherein: a signal conditioning module used for filtering, signal following and differential amplification of the weak heat flow signals to output analog signals; and an embedded processing unit used for converting the analog signals into digital signals and transmitting the digital signals through a communication interface. The signal conditioning module comprises: a filtering circuit used for filtering high-frequency common-mode noise and differential noise in the weak heat flow signals; a signal following circuit connected with the output end of the filtering circuit and used for improving signal driving capacity; and a differential amplification circuit connected with the output end of the signal following circuit and used for amplifying signals.
4. The intelligent monitoring system based on heat flow multi-dimensional features of claim 1, wherein: The measurement mode of the heat flow detection unit comprises a contact measurement mode or a radiation measurement mode; The contact measurement mode is realized by direct contact of a flexible or rigid substrate sensor with a measured object surface; The radiation measurement mode is realized by keeping a preset distance between a sensor and a measured object to collect thermal radiation energy emitted by the target object in a non-contact mode.
5. The intelligent monitoring system based on multi-dimensional characteristics of heat flow according to claim 4, characterized in that: The application further relates to a heat flow detection method, comprising the following steps: collecting heat flow density data of multiple points in a monitoring area through a distributed heat flow sensor array; converting the collected weak heat flow signals into digital signals or analog signals with high reliability through an intelligent transmission circuit; calculating time-domain dynamic features dQ / dt, spatial distribution features DeltaQ and steady-state features |Q| of the digital signals or analog signals with high reliability in real time; 6. The intelligent monitoring system based on multi-dimensional features of heat flow according to claim 1, characterized in that: adaptively adjusting dynamic weight coefficients alpha, beta and gamma for constructing a comprehensive risk index model according to the type of a target application scene; calculating a comprehensive risk index based on the dynamic weight coefficients and the time-domain dynamic features, the spatial distribution features and the steady-state features; and triggering a graded response strategy when the comprehensive risk index exceeds a preset threshold.
7. An intelligent regulation method based on the multi-dimensional characteristics of heat flow, characterized in that: 8.The method of claim 7, wherein the method further comprises: Before calculating the comprehensive risk index, a step of normalizing the time-domain dynamic characteristics, the space-domain distribution characteristics and the steady-state characteristics is further included. 9.The method of claim 7, wherein: The step of adaptively adjusting the dynamic weight coefficients according to the type of the target application scenario comprises: when it is judged that the application scenario is a transient risk scenario, increasing the weight coefficient α corresponding to the time-domain dynamic characteristics; when it is judged that the application scenario is a device local overheating scenario, increasing the weight coefficient γ corresponding to the space-domain distribution characteristics; when it is judged that the application scenario is a long-term stable operation scenario, increasing the weight coefficient β corresponding to the steady-state characteristics. 10.The method of claim 7, wherein the method further comprises: The method trains and optimizes the dynamic weight coefficients α, β and γ by using historical monitoring data and feedback of regulation and control effects through a machine learning algorithm.