An Adaptive Adjustment System and Method for Intelligent Switch Panels Based on Environmental Perception
By monitoring the load status in the smart switch panel and using dynamic compensation and self-learning methods, the onboard temperature is decoupled from the ambient temperature, which solves the problems of misjudgment and lag in the environmental perception and adaptive adjustment of the smart switch panel, and improves the response accuracy and control reliability.
Patent Information
- Application Number
- CN202611523333.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-29
- Publication Date
- 2026-11-17
AI Technical Summary
Existing smart switch panels suffer from discrepancies between onboard temperature and actual ambient temperature in environmental sensing and adaptive adjustment, leading to misjudgments and delayed adjustments, making it difficult to adapt to thermal variations under different installation environments and load conditions.
By monitoring the conduction state of the target load, the initial reference temperature is extracted, and the real-time onboard temperature, current and voltage sequences are obtained synchronously in the dynamic compensation cycle. The theoretical self-heating temperature rise is calculated, and the ambient temperature is decoupled and compared with the target temperature range to generate adjustment commands. At the same time, after the load is disconnected, the thermal environment is self-learned to update the thermal resistance coefficient and time constant.
It improves the accuracy of the smart switch panel's response to the indoor thermal environment and the reliability of its adaptive control, reduces the error in environmental perception caused by high-power loads and long-term conduction, and realizes the panel's gradual adaptation capability.
Smart Images

Figure CN123113391A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of regulation and analysis technology, and in particular to an adaptive regulation system and method for an intelligent switch panel based on environmental perception. Background Technology
[0002] As a crucial entry point for indoor power control and scene linkage, smart switch panels handle the status control and information exchange functions of equipment such as lighting, air conditioning, and ventilation. Their operational stability directly impacts the comfort of the living environment and the efficiency of power management. Meanwhile, indoor ambient temperature is a vital data point affecting human comfort and equipment adjustment strategies; accurately acquiring ambient temperature is essential for achieving refined, low-interference intelligent control.
[0003] It is worth noting that smart switch panels are typically installed on walls and are constantly connected to loads. After the load is turned on, the internal circuitry, terminals, and power devices of the panel generate self-heating, causing a deviation between the onboard temperature and the actual ambient temperature. This deviation is not constant but varies with load power, conduction time, installation location, and wall heat dissipation conditions. Therefore, while the onboard temperature can reflect the surrounding thermal environment, the self-heating component can affect the accuracy of environmental perception, thus impacting the basis for subsequent linkage control decisions.
[0004] However, existing smart switch panels still have shortcomings in environmental perception and adaptive adjustment. Some solutions directly use the onboard temperature as the indoor ambient temperature, which is prone to misjudgment under high-power loads or long-term on-time scenarios; some solutions only use fixed temperature compensation values, which are difficult to adapt to the differences in thermal changes under different installation environments and load conditions; and some solutions focus on simple threshold comparisons, lacking correlation analysis of the temperature rise process caused by load operation and the heat dissipation process after shutdown, resulting in adjustment commands that are prone to lag or deviate from actual environmental requirements, thus limiting the environmental perception accuracy and adaptive control effect of smart switch panels in complex home scenarios. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this application provides an adaptive adjustment system and method for an intelligent switch panel based on environmental perception.
[0006] In a first aspect, this application provides an adaptive adjustment method for an intelligent switch panel based on environmental perception, comprising the following steps: When the target load on the monitoring access panel switches from disconnected to on state, the onboard temperature data collected at the moment of on-state is extracted as the initial reference temperature. Enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The real-time load current and real-time load voltage at the same sampling time are multiplied to generate real-time active power data. The equivalent thermal resistance coefficient and thermal time constant stored locally are read, and the theoretical self-heating temperature rise is calculated recursively according to the change of real-time active power data at adjacent sampling times. Subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupled ambient temperature. The decoupled ambient temperature is compared with the target temperature range, and an adjustment command matching the comparison result is generated. When the target load is switched from on to off, the dynamic compensation loop is exited and the thermal environment self-learning loop is entered. The cooling temperature sequence is obtained in the thermal environment self-learning loop, the temperature decay characteristics are extracted from the cooling temperature sequence, and the equivalent thermal resistance coefficient and thermal time constant are updated according to the temperature decay characteristics.
[0007] Secondly, this application provides an environmentally aware intelligent switch panel adaptive adjustment system, including: The panel monitoring module is used to monitor the target load connected to the panel as it switches from disconnected to on state, and extracts the onboard temperature data collected at the moment of on-state as the initial reference temperature. The compensation loop module is used to enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The recursive calculation module is used to multiply the real-time load current and real-time load voltage at the same sampling time to generate real-time active power data, read the equivalent thermal resistance coefficient and thermal time constant stored locally, and recursively calculate the theoretical self-heating temperature rise value according to the change of real-time active power data at adjacent sampling times. The decoupling calculation module is used to subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupling ambient temperature. The instruction generation module is used to compare the decoupled ambient temperature with the target temperature range and generate adjustment instructions that match the comparison results. The self-learning loop module is used to monitor the target load switching from the on to the off state, exit the dynamic compensation loop and enter the thermal environment self-learning loop. The extraction and update module is used to obtain the cooling temperature sequence in the thermal environment self-learning loop, extract the temperature decay characteristics from the cooling temperature sequence, and update the equivalent thermal resistance coefficient and thermal time constant according to the temperature decay characteristics.
[0008] In summary, this application includes at least one of the following beneficial technical effects: 1. This application provides an adaptive adjustment method for intelligent switch panels based on environmental perception. By monitoring the target load connected to the panel as it switches from disconnected to on, and extracting the onboard temperature data collected at the moment of on-state as the initial reference temperature, the intelligent switch panel establishes an environmental temperature reference when the load has not yet generated a significant self-heating effect. Subsequently, in the dynamic compensation loop, the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence are acquired synchronously at the sampling time. Real-time active power data is generated from the real-time load current and real-time load voltage at the same sampling time. The theoretical self-heating temperature rise value is then calculated by recursively combining the locally stored equivalent thermal resistance coefficient and thermal time constant, thereby separating the heat change of the panel itself due to the load on-state from the onboard temperature data. The above processing method no longer simply relies on fixed compensation values or directly uses the onboard temperature as the ambient temperature. Instead, it associates load power changes, thermal response parameters and temperature sampling time sequence into a continuous data processing chain, so that the theoretical self-heating temperature rise value can change dynamically with the actual operating state of the target load. This improves the authenticity and stability of the decoupled ambient temperature and effectively reduces the errors caused by high-power loads, long-term conduction and voltage and current fluctuations on the environmental perception results. 2. By obtaining the decoupled ambient temperature and comparing it with the target temperature range to generate matching adjustment commands, the adjustment judgment is based on environmental data after deducting the panel's self-heating interference, thereby improving the accuracy of the smart switch panel's response to changes in the indoor thermal environment. Simultaneously, after detecting that the target load has switched from an on to an off state, the dynamic compensation loop exits and enters a thermal environment self-learning loop. Temperature decay characteristics are extracted through the cooling temperature sequence, and the equivalent thermal resistance coefficient and thermal time constant are updated accordingly. This allows the next round of theoretical self-heating temperature rise calculation to absorb changes in the current installation environment, wall heat dissipation conditions, and load operating characteristics. Thus, the dynamic compensation loop and the thermal environment self-learning loop form a closed-loop synergistic relationship: the former eliminates the interference of load self-heating on ambient temperature judgment in real time, while the latter uses the natural cooling process after shutdown to reverse-correct the compensation parameters. The two support each other, enabling the smart switch panel to gradually adapt to different usage scenarios, avoiding the problems of fixed thresholds, rigid compensation, and lag in traditional solutions, and improving environmental perception accuracy and adaptive control reliability. Attached Figure Description
[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0010] Figure 1This is a flowchart of a method for adaptive adjustment of an intelligent switch panel based on environmental perception, according to an embodiment of this application.
[0011] Figure 2 This is a schematic diagram of an intelligent switch panel adaptively adjusting based on environmental perception, according to an embodiment of this application. Detailed Implementation
[0012] The following description, in conjunction with the implementation of the present invention, is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention, and all such modifications and additions should fall within the protection scope of the present invention.
[0013] Example 1
[0014] This application applies to the adaptive adjustment of intelligent switch panels in indoor environmental control scenarios. Intelligent switch panels typically have onboard temperature acquisition capabilities. However, after the target load is turned on, the load current generates heat as it passes through terminals, relays, power devices, and internal panel wiring, causing the real-time onboard temperature to no longer reflect the actual ambient temperature. The core of this implementation is to process the target load's on-state, real-time load current sequence, real-time load voltage sequence, and real-time onboard temperature sequence at the same sampling time. First, the theoretical self-heating temperature rise caused by the target load's operation is estimated. Then, this temperature rise is subtracted from the real-time onboard temperature to obtain a decoupled ambient temperature that more closely approximates the indoor thermal environment. Furthermore, during the cooling process after the target load is turned off, the equivalent thermal resistance coefficient and thermal time constant are reversed, making the temperature rise calculation in subsequent dynamic compensation cycles more consistent with the current installation environment.
[0015] This application discloses an adaptive adjustment method for an intelligent switch panel based on environmental perception.
[0016] Reference Figure 1 An adaptive adjustment method for intelligent switch panels based on environmental perception includes the following steps: When the target load on the monitoring access panel switches from disconnected to on state, the onboard temperature data collected at the moment of on-state is extracted as the initial reference temperature. Furthermore, when the target load on the monitoring access panel switches from disconnected to on state, the onboard temperature data collected at the moment of on-state is extracted as the initial reference temperature, including: The real-time load current sequence of the target load is obtained according to a fixed sampling frequency; Divide the real-time load current sequence into adjacent pre-time windows and post-time windows, and calculate the mean pre-time current within the pre-time window and the mean post-time current within the post-time window respectively. The change in current is obtained by subtracting the average value of the preceding current from the average value of the subsequent current. When the change in current is greater than the preset jump threshold, the target load is determined to switch from the disconnected state to the on state, and the starting point of the subsequent time window is marked as the state switching time. Extract onboard temperature data within a preset time period before the state switching time and combine them into an environmental baseline sequence; The highest and lowest temperature data in the environmental reference sequence are removed, and the arithmetic mean of the remaining onboard temperature data in the environmental reference sequence is calculated to generate the initial reference temperature.
[0017] In one specific implementation, the processor within the smart switch panel executes an environmentally aware adaptive adjustment method. The processor is connected to onboard temperature, current, and voltage acquisition circuits. The onboard temperature acquisition circuit outputs onboard temperature data in degrees Celsius; the current acquisition circuit outputs load current data in amperes; and the voltage acquisition circuit outputs load voltage data in volts. The processor binds and stores this data according to timestamps, with timestamp accuracy down to the millisecond level, although in actual products, it can also be saved down to the second level. The target load is the electrical load connected to the panel and controlled by the smart switch panel, such as lighting loads, ventilation loads, heating loads, and other loads related to indoor environmental regulation.
[0018] During method execution, the processor first monitors the target load on the connected panel to switch from an off to an on state. When the target load is not yet on, the self-heating generated inside the panel by the target load is relatively small, making the onboard temperature data collected at this time more suitable as the initial reference for ambient temperature. Therefore, once the target load is detected to have switched from off to on, the processor extracts the onboard temperature data collected at the moment of on-state as the initial reference temperature. The unit of the initial reference temperature is the same as the onboard temperature data, both in degrees Celsius, and its value is, for example, 25.8°C. This initial reference temperature then serves as the starting point for the environment in the dynamic compensation loop, allowing subsequent temperature rises in the real-time onboard temperature sequence to be distinguished from the self-heating temperature rise generated by the target load's operation.
[0019] Specifically, monitoring the target load's transition from an off to a on state can be achieved using a window comparison method with a real-time load current sequence. The processor acquires the real-time load current sequence of the target load at a fixed sampling frequency, which can range from 10Hz to 100Hz. For example, at 20Hz, load current data is collected every 50 milliseconds. Each load current data point in the real-time load current sequence is marked with a collection time stamp. The processor divides the real-time load current sequence into adjacent pre-switching and post-switching time windows. The pre-switching time window is located before the candidate switching point, and the post-switching time window is located after the candidate switching point. Both windows can have the same length, for example, 0.5 seconds. The processor calculates the average pre-switching current within the pre-switching time window and the average post-switching current within the post-switching time window, respectively. The average pre-switching current reflects the current background before the target load switches, and the average post-switching current reflects the load current level after the candidate switching point.
[0020] The processor calculates the current change by subtracting the average current from the average current before. When the current change exceeds a preset threshold, the target load is determined to switch from an off to a on state, and the start point of the post-time window is marked as the state transition moment. The preset threshold can be determined based on the minimum operating current of the target load; for example, it can be 0.03A to 0.08A for lighting loads with low rated current, and 0.1A to 0.3A for environmental regulation loads with high power. The threshold does not need to cover a uniform value for all products; it can be written at the factory according to the load type, or it can be determined by installation and commissioning data. Using the difference between the average current before and after the current can filter out false triggers caused by single-point sampling noise, making the state transition moment more stable.
[0021] Once the state transition time is determined, the processor extracts the onboard temperature data within a preset time period prior to the state transition time and combines it into an environmental reference sequence. The preset time period can be 5 to 60 seconds, but is typically 10 to 30 seconds in scenarios with relatively gradual room temperature changes. The onboard temperature data in the environmental reference sequence comes from before the target load is turned on, at which point the target load has not yet formed significant self-heating, making it more representative of the environmental starting point. To reduce the impact of instantaneous sampling spikes, the processor removes the highest and lowest temperature data from the environmental reference sequence and calculates the arithmetic mean of the remaining onboard temperature data to generate the initial reference temperature. For example, if the environmental reference sequence contains 20 onboard temperature data points, after removing the maximum and minimum values, the average of the remaining 18 onboard temperature data points is taken to obtain the initial reference temperature. This process does not change the temperature dimension; the generated initial reference temperature is still in degrees Celsius and is written to the local cache as the temperature reference at the start of the dynamic compensation loop.
[0022] Enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. Furthermore, a dynamic compensation loop is entered, synchronously acquiring real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time, including: Using the state transition time as the starting point of the loop, a sampling time sequence is generated according to the compensation sampling interval; The onboard temperature data, load current data, and load voltage data collected at each sampling moment are bound into a single frame of operating condition data; The single-frame operating condition data is sorted according to the sampling time sequence to obtain the operating condition data sequence; In the operating condition data sequence, abnormal frames with time intervals greater than the compensation sampling interval are identified, and the onboard temperature data, load current data and load voltage data corresponding to the abnormal frames are supplemented by the average of the same type of data in adjacent valid frames. The real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence were extracted from the completed operating condition data sequence.
[0023] In one specific implementation, after the target load is confirmed to be on, the processor enters a dynamic compensation loop and synchronously acquires the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The sampling time can be generated by the processor's internal timer, for example, one sampling time is generated every 1 second, or a compensated sampling interval between 0.5 seconds and 2 seconds can be used in high-power load scenarios. At each sampling time, the onboard temperature acquisition circuit outputs a real-time onboard temperature, the current acquisition circuit outputs a real-time load current, and the voltage acquisition circuit outputs a real-time load voltage. The processor associates the three with the same sampling time to avoid misalignment between temperature data and power data. In actual storage, the real-time onboard temperature sequence can be represented as a temperature array arranged by time, the real-time load current sequence can be represented as a current array arranged by time, and the real-time load voltage sequence can be represented as a voltage array arranged by time, with each array having the same sampling time index.
[0024] Specifically, given that the aforementioned state transition moment has already been established, upon entering the dynamic compensation loop, the processor uses the state transition moment as the loop starting point and generates a sampling time sequence according to the compensation sampling interval. The compensation sampling interval can differ from the fixed sampling frequency. The fixed sampling frequency is geared towards accurately identifying the on / off moment, while the compensation sampling interval is geared towards subsequent stable calculations of temperature and power. For example, the fixed sampling frequency might be 20Hz, while the compensation sampling interval is 1 second. The sampling time sequence is formed by incrementing the compensation sampling interval from the state transition moment, recording the time position of each data processing step in the dynamic compensation loop.
[0025] At each sampling moment, the processor binds the acquired onboard temperature data, load current data, and load voltage data into a single frame of operating condition data. A single frame of operating condition data includes at least four items: sampling moment, onboard temperature data, load current data, and load voltage data, recorded as follows: and The processor sorts the single-frame operating condition data according to the sampling time sequence to obtain the operating condition data sequence. If the time interval between two adjacent frames in the operating condition data sequence is greater than the compensation sampling interval, it indicates the existence of an abnormal frame. Abnormal frames may originate from sampling delay, communication jitter, or short-term missing readings. The processor uses the average of similar data in adjacent valid frames to fill in the onboard temperature data, load current data, and load voltage data corresponding to the abnormal frame. For example, the onboard temperature data of the missing frame is obtained by averaging the onboard temperature data of the previous valid frame and the onboard temperature data of the next valid frame. The load current data and load voltage data are also averaged separately according to similar data. In this process, temperature is only averaged with temperature, current is only averaged with current, and voltage is only averaged with voltage, avoiding dimensional errors caused by mixing different physical quantities in the calculation. The completed operating condition data sequence is then split, and the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence are extracted from it for power calculation and temperature rise recursion.
[0026] The real-time load current and real-time load voltage at the same sampling time are multiplied to generate real-time active power data. The equivalent thermal resistance coefficient and thermal time constant stored locally are read, and the theoretical self-heating temperature rise is calculated recursively according to the change of real-time active power data at adjacent sampling times. Furthermore, the real-time load current and real-time load voltage at the same sampling time are multiplied to generate real-time active power data, including: The real-time load current and real-time load voltage at the same sampling time are multiplied to generate single-point power data. The single-point power data are arranged in chronological order of sampling time to form a real-time active power sequence; Calculate the power mean of the real-time active power sequence within a predetermined number of bits before the current sampling time to generate smoothed active power data; The smoothed active power data is matched with the locally stored power segmentation table to determine the power segment corresponding to the target load. Read the power correction coefficient corresponding to the power segment, and multiply the power correction coefficient by the equivalent thermal resistance coefficient to obtain the segment equivalent thermal resistance coefficient.
[0027] Furthermore, the theoretical self-heating temperature rise is calculated recursively based on the changes in real-time active power data at adjacent sampling times, including: Extract the smoothed active power data at the current sampling time and the theoretical self-heating temperature rise value at the previous sampling time; Based on smoothed active power data, the time interval between adjacent sampling times, and the piecewise equivalent thermal resistance coefficient, the heat generation increment at the current sampling time is calculated. Based on the theoretical self-heating temperature rise at the previous sampling time, the time interval between adjacent sampling times, and the thermal time constant, the heat dissipation attenuation at the current sampling time is calculated. Add the heat generation increment to the theoretical self-heating temperature rise at the previous sampling time, and subtract the heat dissipation attenuation to obtain the theoretical self-heating temperature rise at the current sampling time. Write the theoretical self-heating temperature rise value at the current sampling time into the self-heating temperature rise sequence.
[0028] In one specific implementation, within the dynamic compensation loop, the processor multiplies the real-time load current and real-time load voltage at the same sampling time to generate real-time active power data. If the current sampling time is denoted as... The real-time load current is denoted as The real-time load voltage is denoted as Real-time active power data Press Obtained. Among them, Indicates the target load at the sampling time The current, measured in amperes; This represents the voltage of the target load at the same sampling time, in volts. This represents the power consumed by the target load at the sampling moment, in watts. Since watts are obtained by multiplying amperes by volts, this calculation avoids dimensional inconsistencies. The processor then reads the locally stored equivalent thermal resistivity and thermal time constant. The equivalent thermal resistivity characterizes the steady-state temperature rise caused by unit power, and its unit can be ℃ / W; the thermal time constant characterizes the speed at which the panel goes from heating up to near thermal equilibrium, and its unit can be seconds. Both can be written during factory calibration or updated after a thermal environment self-learning cycle.
[0029] The recursive calculation of the theoretical self-heating temperature rise is related to the changes in real-time active power data at adjacent sampling times. If the theoretical self-heating temperature rise at the previous sampling time is... The smoothed active power data at the current sampling time is The segmented equivalent thermal resistance coefficient is The time interval between adjacent sampling times is The thermal time constant is Then the theoretical self-heating temperature rise at the current sampling time is... Press Perform a recursive calculation. In the formula, This represents the steady-state temperature rise under the current power conditions, in degrees Celsius. This is a dimensionless proportion, representing the percentage of the current sampling interval in the thermal response process; This represents the heat increment at the current sampling time; This represents the heat dissipation attenuation at the current sampling moment. This recursive relationship reflects the physical process by which the panel's self-heating gradually accumulates from low to high and is offset by the heat dissipation process. The resulting theoretical self-heating temperature rise is still in degrees Celsius and can be directly compared with the real-time onboard temperature.
[0030] Specifically, for generating real-time active power data, the processor multiplies the real-time load current and real-time load voltage at each sampling time to generate single-point power data. Subsequently, the processor arranges the single-point power data in the order of sampling times to form a real-time active power sequence. In the initial stage of conduction, the actual load may experience current surges, and voltage sampling may also have slight fluctuations. If each single-point power data point is directly used for temperature rise estimation, the theoretical self-heating temperature rise value will be affected by transient disturbances. Therefore, the processor calculates the power average of the real-time active power sequence within a predetermined number of positions before the current sampling time to generate smoothed active power data. The predetermined number of positions can be 3 to 10. For example, when it is 5, the smoothed active power data at the current sampling time is the arithmetic mean of the single-point power data at the current sampling time and the four sampling times preceding it. If the predetermined number of positions is insufficient in the initial stage of conduction, the power average is calculated based on the number of single-point power data points already obtained, ensuring that the dynamic compensation cycle can start immediately after the state transition.
[0031] After the smoothed active power data is generated, the processor matches it with the locally stored power segmentation table to determine the power segment corresponding to the target load. The power segmentation table can be configured according to product type, for example, 0W to 50W for low power segments, 50W to 300W for medium power segments, and above 300W for high power segments; it can also be adjusted according to on-site calibration data. Under different power segments, the contributions of heat-generating locations and heat dissipation paths within the panel are not entirely consistent. Therefore, the processor reads the power correction coefficient corresponding to the power segment and multiplies it by the equivalent thermal resistance coefficient to obtain the segmented equivalent thermal resistance coefficient. The power correction coefficient is dimensionless, while the equivalent thermal resistance coefficient is in °C / W. After multiplying, the unit of the segmented equivalent thermal resistance coefficient remains °C / W. This segmentation process allows the same panel to use different thermal resistance correction results under different target load power conditions, avoiding the use of a single equivalent thermal resistance coefficient to cover all operating conditions.
[0032] In the recursive process of theoretical self-heating temperature rise, the processor extracts the smoothed active power data at the current sampling moment and the theoretical self-heating temperature rise at the previous sampling moment. Upon initial entry into the dynamic compensation loop, the theoretical self-heating temperature rise at the previous sampling moment can be set to 0℃ because the target load is disconnected before the state transition, and the self-heating corresponding to the target load has not yet begun to accumulate. The processor calculates the heat increment at the current sampling moment based on the smoothed active power data, the time interval between adjacent sampling moments, and the piecewise equivalent thermal resistance coefficient; simultaneously, it calculates the heat dissipation attenuation at the current sampling moment based on the theoretical self-heating temperature rise at the previous sampling moment, the time interval between adjacent sampling moments, and the thermal time constant. Using the aforementioned... In the recursive relationship, both the heat increment and the heat dissipation decay are in degrees Celsius. They represent the contribution of the target load power input to the temperature rise and the contribution of the temperature rise reduction caused by the heat released by the panel to the surrounding environment within the current sampling interval, respectively. The processor adds the heat increment to the theoretical self-heating temperature rise value of the previous sampling moment and subtracts the heat dissipation decay to obtain the theoretical self-heating temperature rise value of the current sampling moment. Then, the theoretical self-heating temperature rise value of the current sampling moment is written into the self-heating temperature rise sequence. The self-heating temperature rise sequence corresponds one-to-one with the sampling moment sequence, and subsequent decoupled ambient temperature calculations directly read the data at the same sampling moment.
[0033] Subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupled ambient temperature. Furthermore, the theoretical self-heating temperature rise is subtracted from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupled ambient temperature, including: Extract the real-time onboard temperature at the current sampling moment from the real-time onboard temperature sequence; Extract the theoretical self-heating temperature rise value at the current sampling moment from the self-heating temperature rise sequence; The single-point decoupled ambient temperature is generated by subtracting the theoretical self-heating temperature rise at the current sampling time from the real-time onboard temperature at the current sampling time. The single-point decoupled ambient temperatures are arranged in chronological order of sampling time to form a decoupled ambient temperature sequence; Calculate the difference between adjacent single-point decoupled ambient temperatures in the decoupled ambient temperature sequence to obtain the ambient temperature difference sequence; When the positive and negative directions of the differences in the predetermined number of positions in the ambient temperature difference sequence are the same and the cumulative difference is greater than the preset temperature threshold, the cumulative difference is incorporated into the initial reference temperature to generate a corrected reference temperature.
[0034] In one specific implementation, after obtaining the theoretical self-heating temperature rise, the processor extracts the real-time onboard temperature corresponding to the sampling time from the real-time onboard temperature sequence, and subtracts the theoretical self-heating temperature rise from this real-time onboard temperature to obtain the decoupled ambient temperature. For example, if the real-time onboard temperature at a certain sampling time is 30.6℃ and the theoretical self-heating temperature rise is 3.1℃, then the corresponding decoupled ambient temperature is 27.5℃. This subtraction is performed under the same temperature dimensions, and the result represents the estimated ambient temperature after deducting the thermal influence of the target load. Compared to directly using the real-time onboard temperature, the decoupled ambient temperature reduces the interference of target load power changes on temperature judgment, making it less likely that the adjustment command will deviate from the actual indoor thermal environment due to the panel's own temperature rise.
[0035] Specifically, in the process of forming the decoupled ambient temperature, the processor extracts the real-time onboard temperature at the current sampling moment from the real-time onboard temperature sequence and the theoretical self-heating temperature rise value at the current sampling moment from the self-heating temperature rise sequence. Both sets of data have the same sampling moment index, so there is no time mismatch in the difference processing. The processor subtracts the theoretical self-heating temperature rise value at the current sampling moment from the real-time onboard temperature at the current sampling moment to generate a single-point decoupled ambient temperature, and arranges the single-point decoupled ambient temperatures in chronological order of sampling moment to form a decoupled ambient temperature sequence. Since the deduction of the target load self-heating is performed point by point, the decoupled ambient temperature sequence retains the information of the indoor environment changing over time and weakens the superimposed influence of the heat source inside the panel on the onboard temperature data.
[0036] In practical applications, the indoor environment itself may also change slowly. For example, opening windows, starting air conditioning, and sunlight shining on the wall will all change the temperature near the wall. To avoid introducing deviation by keeping the initial reference temperature constant throughout the conduction period, the processor calculates the difference between adjacent single-point decoupled ambient temperatures in the decoupled ambient temperature sequence, resulting in an ambient temperature difference sequence. Each difference in the ambient temperature difference sequence is still in degrees Celsius, representing the amount of change in decoupled ambient temperature at adjacent sampling times. When the differences of a predetermined number of digits in the ambient temperature difference sequence are in the same positive or negative direction and the cumulative difference is greater than a preset temperature threshold, the processor incorporates the cumulative difference into the initial reference temperature to generate a corrected reference temperature. The predetermined number of digits can be 3 to 8, and the preset temperature threshold can be 0.3℃ to 1.0℃, which can be determined according to the resolution of the onboard temperature acquisition circuit and the rate of change of indoor temperature. If multiple consecutive differences are positive and the cumulative amount exceeds the preset temperature threshold, it indicates that the decoupled ambient temperature sequence is rising in the same direction; if multiple consecutive differences are negative and the absolute change of the cumulative amount meets the requirements, it indicates that the ambient temperature is decreasing in the same direction. After the cumulative difference is incorporated into the initial reference temperature, the corrected reference temperature can reflect the changes in the environmental reference during conduction, avoiding the use of the instantaneous temperature during conduction to represent the entire conduction process.
[0037] The decoupled ambient temperature is compared with the target temperature range, and an adjustment command matching the comparison result is generated. Furthermore, the decoupled ambient temperature is compared with the target temperature range to generate adjustment commands that match the comparison results, including: Read the lower and upper temperature limits of the target temperature range; Subtract the upper temperature limit from the single-point decoupled ambient temperature at the current sampling time to obtain the upper limit deviation data; Subtract the single-point decoupled ambient temperature at the current sampling time from the lower limit value to obtain the lower limit deviation data; When the upper limit deviation data is greater than zero, a cooling adjustment command is generated; When the lower limit deviation data is greater than zero, a temperature-increasing adjustment command is generated; When both the upper limit deviation data and the lower limit deviation data are not greater than zero, a hold-type adjustment command is generated; The adjustment command is combined with the current sampling time, the single-point decoupled ambient temperature, and the corrected reference temperature to form an adjustment record.
[0038] In one specific implementation, the processor then compares the decoupled ambient temperature with the target temperature range and generates an adjustment command matching the comparison result. The target temperature range can be pre-stored in the smart switch panel or set by the user and written to the local storage area, for example, a lower temperature limit of 24°C and an upper temperature limit of 27°C. When the decoupled ambient temperature at the current sampling time is higher than the upper temperature limit, the processor generates a cooling adjustment command; when it is lower than the lower temperature limit, it generates a heating adjustment command; and when it is between the lower and upper temperature limits, it generates a holding adjustment command. The adjustment commands form an adjustment record locally and can be used for the smart switch panel's own scene linkage strategy, such as adjusting the working level of associated loads, recording environmental control status, or for local control logic to read.
[0039] Specifically, when generating adjustment instructions, the processor reads the lower and upper temperature limits of the target temperature range. The target temperature range can be configured according to room type; for example, the bedroom's nighttime control range is set to 24℃ to 26℃, and the living room's daytime control range is set to 25℃ to 27℃. The processor subtracts the upper temperature limit from the current sampling time's single-point decoupled ambient temperature to obtain the upper limit deviation data; simultaneously, it subtracts the current sampling time's single-point decoupled ambient temperature from the lower temperature limit to obtain the lower limit deviation data. When the upper limit deviation data is greater than zero, it indicates that the current sampling time's single-point decoupled ambient temperature is higher than the upper temperature limit, and the processor generates a cooling adjustment instruction. When the lower limit deviation data is greater than zero, it indicates that the current sampling time's single-point decoupled ambient temperature is lower than the lower temperature limit, and the processor generates a heating adjustment instruction. If both the upper and lower limit deviation data are not greater than zero, the current sampling time's single-point decoupled ambient temperature is within the target temperature range, and the processor generates a hold adjustment instruction. After each adjustment instruction is generated, the processor combines the adjustment instruction with the current sampling time, the single-point decoupled ambient temperature, and the correction reference temperature to form an adjustment record. The adjustment records can be written to the local cache in chronological order, which makes it easier to determine whether the adjustment instructions switch frequently in the same scenario, and can also provide the environmental state background during the conduction period for the thermal environment self-learning loop.
[0040] When the target load is switched from on to off, the dynamic compensation loop is exited and the thermal environment self-learning loop is entered. The cooling temperature sequence is obtained in the thermal environment self-learning loop, the temperature decay characteristics are extracted from the cooling temperature sequence, and the equivalent thermal resistance coefficient and thermal time constant are updated according to the temperature decay characteristics.
[0041] Furthermore, the monitoring target load switches from an on to an off state, exits the dynamic compensation loop, and enters the thermal environment self-learning loop, including: Obtain the real-time load current sequence of the target load in the dynamic compensation loop; Divide the real-time load current sequence into adjacent conduction end windows and disconnection confirmation windows, and calculate the average end current in the conduction end window and the average confirmation current in the disconnection confirmation window respectively. The change in turn-off current is obtained by subtracting the average value of the confirmed current from the average value of the end current. When the change in turn-off current is greater than the preset turn-off threshold, the target load is determined to switch from the on state to the off state. Mark the start point of the disconnect confirmation window as the disconnection time; Read the real-time onboard temperature, theoretical self-heating temperature rise, and decoupled ambient temperature at the time of shutdown, and generate an initial cooling record. The thermal environment self-learning loop is initiated based on the initial cooling record.
[0042] Furthermore, the cooling temperature sequence is obtained during the thermal environment self-learning loop, and temperature decay features are extracted from the cooling temperature sequence, including: Using the shutdown moment as the starting point for cooling sampling, onboard temperature data is continuously collected according to the cooling sampling interval to form a cooling temperature sequence. Calculate the temperature difference between adjacent onboard temperature data in the cooling temperature sequence to obtain the cooling difference sequence; When the absolute value of the temperature difference of a predetermined number in the cooling difference sequence is less than the cooling stability threshold, the average value of the corresponding onboard temperature data is calculated to generate the cooling endpoint temperature. Subtract the cooling endpoint temperature from each onboard temperature data in the cooling temperature sequence to obtain the cooling residual temperature sequence; Divide each cooling residual temperature data in the cooling residual temperature sequence by the theoretical self-heating temperature rise value corresponding to the turn-off time to obtain the normalized decay sequence. The normalized decay sequence is fitted with exponential decay to extract temperature decay features.
[0043] Furthermore, the equivalent thermal resistivity and thermal time constant are updated based on the temperature decay characteristics, including: The fitted thermal time constant is determined based on the temperature decay characteristics; Read the real-time load current sequence and real-time load voltage sequence arranged according to the sampling time during this conduction period; Multiply the real-time load current and real-time load voltage at the same sampling time to generate a conduction power sequence; Calculate the arithmetic mean of the conduction power sequence to generate the average conduction power data; Divide the theoretical self-heating temperature rise corresponding to the turn-off time by the average conduction power data to generate the fitted equivalent thermal resistance coefficient. Calculate the time constant deviation between the fitted thermal time constant and the locally stored thermal time constant, and calculate the thermal resistance deviation between the fitted equivalent thermal resistance coefficient and the locally stored equivalent thermal resistance coefficient. When both the time constant deviation and the thermal resistance deviation fall within the update range, the updated thermal time constant and the updated equivalent thermal resistance coefficient are generated by weighted averaging. Update the thermal time constant, update the equivalent thermal resistance coefficient, and write the average conduction power data to the local parameter table.
[0044] In one specific implementation, during the continuous execution of the dynamic compensation loop, the processor also monitors the target load as it switches from an on to an off state. Once the target load is disconnected, the real-time load current decreases, and the heat input caused by the target load disappears, causing the panel to transition from the heating phase to the cooling phase. After determining that the target load has switched from an on to an off state, the processor exits the dynamic compensation loop and enters the thermal environment self-learning loop. The thermal environment self-learning loop no longer estimates heat generation based on power input, but instead collects the cooling temperature sequence after disconnection, extracts temperature decay characteristics from the cooling temperature sequence, and updates the equivalent thermal resistance coefficient and thermal time constant based on the temperature decay characteristics. In this way, after each complete on-off cycle of the target load, a cooling data related to the current wall, junction box, load power, and indoor ventilation conditions is left behind. The processor corrects its local parameters accordingly, so that the next dynamic compensation loop does not rely on fixed compensation values.
[0045] Specifically, the monitoring of the target load switching from a conducting to a disconnected state is also based on the real-time load current sequence. In the dynamic compensation loop, the processor continuously acquires the real-time load current sequence of the target load and divides it into adjacent conducting end windows and disconnection confirmation windows. The conducting end window is located before the candidate turn-off point, and the disconnection confirmation window is located after the candidate turn-off point. The window length can be the same as the conducting identification stage, or it can be 0.3 seconds to 1 second depending on the load turn-off characteristics. The processor calculates the average end current within the conducting end window and the average confirmation current within the disconnection confirmation window, and then subtracts the average confirmation current from the average end current to obtain the change in turn-off current. When the change in turn-off current exceeds a preset turn-off threshold, the processor determines that the target load has switched from a conducting to a disconnected state. The preset turn-off threshold can be the same as the preset jump threshold, or it can be set separately considering leakage current, standby current, and sampling noise after the load is disconnected.
[0046] Once the target load is determined to be disconnected, the processor marks the start point of the disconnection confirmation window as the shutdown moment. Subsequently, the processor reads the real-time onboard temperature, theoretical self-heating temperature rise, and decoupled ambient temperature corresponding to the shutdown moment to generate an initial cooling record. In the initial cooling record, the real-time onboard temperature represents the actual temperature of the panel when it enters the cooling phase, the theoretical self-heating temperature rise represents the estimated amount of self-heat accumulated by the target load at that time, and the decoupled ambient temperature represents the estimated ambient temperature after deducting self-heating. These three data points together define the initial state of the cooling process, and the processor initiates a thermal environment self-learning loop based on the initial cooling record.
[0047] In the thermal environment self-learning loop, the processor uses the shutdown moment as the starting point for cooling sampling and continuously collects onboard temperature data according to the cooling sampling interval to form a cooling temperature sequence. The cooling sampling interval can be 1 to 10 seconds. If the target load power is high and the panel temperature rise is large, a shorter interval can be used to retain details of the early part of the cooling curve; if the target load power is low, a longer interval can be used to reduce storage requirements. The onboard temperature data in the cooling temperature sequence gradually approaches the ambient thermal temperature over time. The processor calculates the temperature difference between adjacent onboard temperature data in the cooling temperature sequence to obtain a cooling difference sequence. Since the onboard temperature usually decreases during the cooling process, the temperature difference may be negative. Therefore, the processor uses the absolute value of the temperature difference when determining the steady state. If the absolute value of the temperature difference of a predetermined number of digits in the cooling difference sequence is less than the cooling stability threshold, the cooling temperature sequence is considered to have entered a stage of smaller changes. The number of preset bits can be 5 to 20, and the cooling stability threshold can be 0.02℃ to 0.1℃. The specific value can be determined according to the resolution of the onboard temperature acquisition circuit. For example, when the temperature resolution is 0.1℃, the cooling stability threshold can be 0.1℃; when the temperature resolution is 0.01℃, the cooling stability threshold can be 0.03℃.
[0048] Once the cooling stabilization condition is met, the processor averages the corresponding onboard temperature data to generate the cooling endpoint temperature. The cooling endpoint temperature represents the temperature level at which the panel approaches thermal equilibrium after the target load is disconnected. Subsequently, the processor subtracts the cooling endpoint temperature from each onboard temperature data point in the cooling temperature sequence to obtain the residual cooling temperature sequence. The unit of the residual cooling temperature data is degrees Celsius, representing the temperature rise remaining relative to the cooling endpoint temperature at the corresponding sampling time. The processor then divides each residual cooling temperature data point in the residual cooling temperature sequence by the theoretical self-heating temperature rise corresponding to the shutdown time to obtain the normalized decay sequence. The theoretical self-heating temperature rise corresponding to the shutdown time is also in degrees Celsius; therefore, the normalized decay sequence is dimensionless, representing the proportion of residual temperature relative to the estimated self-heating at the shutdown time. For residual cooling temperature data that is less than zero or close to zero, the processor can remove them before fitting to prevent noise in the stabilization tail from affecting the temperature decay characteristics.
[0049] The processor performs exponential decay fitting on the normalized decay sequence to extract temperature decay features. If the cooling sampling time is denoted as... The shutdown time is recorded as The normalized decay data in the normalized decay sequence is denoted as Then it can be adopted Describe the cooling process, where, This represents the fitted thermal time constant, in seconds. This represents the natural exponential function. To facilitate processor computation, the natural logarithm can be taken for positive normalized decay data, yielding... Then based on the time difference and Linear relationship estimation between The physical significance here is that... The larger the value, the slower the residual heat decays on the panel; The smaller the value, the faster the panel dissipates heat to the surrounding environment. After extracting the temperature decay characteristics from the cooling temperature sequence, the thermal environment self-learning loop obtains the actual heat dissipation characteristics under the current installation conditions.
[0050] During the parameter update phase, the processor determines the fitted thermal time constant based on the temperature decay characteristics. This fitted thermal time constant is derived from the current cooling process and does not directly replace the locally stored thermal time constant; instead, it is first compared with the original parameters to determine the deviation. Simultaneously, the processor reads the real-time load current sequence and real-time load voltage sequence arranged by sampling time during the current conduction period. It multiplies the real-time load current and real-time load voltage at the same sampling time to generate the conduction power sequence. The conduction power sequence covers the target load's operating process from the state switching time to the turn-off time. The processor calculates the arithmetic mean of the conduction power sequence to generate the average conduction power data. The unit of the average conduction power data is watts, reflecting the average level of the target load's thermal input to the panel during the current conduction period.
[0051] Subsequently, the processor divides the theoretical self-heating temperature rise corresponding to the turn-off moment by the average conduction power data to generate a fitted equivalent thermal resistance coefficient. If the theoretical self-heating temperature rise corresponding to the turn-off moment is... The average power during conduction is Then the fitted equivalent thermal resistance coefficient It can be represented as In the formula, The unit is Celsius. The unit is watt. The unit is ℃ / W, representing the temperature rise relationship corresponding to the unit average power under the current conduction and cooling conditions. To avoid parameter abnormalities caused by an excessively small divisor in low-power scenarios, the processor can abandon the current update of the equivalent thermal resistance coefficient when the average conduction power data is lower than the minimum effective power threshold, and only retain the cooling temperature sequence as the running record. The minimum effective power threshold can be set according to the power sampling accuracy, for example, from 5W to 20W.
[0052] After the fitted thermal time constant and the fitted equivalent thermal resistance coefficient are formed, the processor calculates the time constant deviation between the fitted thermal time constant and the locally stored thermal time constant, and calculates the thermal resistance deviation between the fitted equivalent thermal resistance coefficient and the locally stored equivalent thermal resistance coefficient. The time constant deviation can be an absolute value or a relative deviation. In practice, to facilitate consistent judgment across different panel models, a relative deviation is often used. For example, the time constant deviation is the absolute value of the fitted thermal time constant minus the locally stored thermal time constant, divided by the locally stored thermal time constant. The thermal resistance deviation is the absolute value of the fitted equivalent thermal resistance coefficient minus the locally stored equivalent thermal resistance coefficient, divided by the locally stored equivalent thermal resistance coefficient. Both deviations are dimensionless data, facilitating comparison with the update range. The update range can be 0% to 30%, or a narrower range can be set according to product calibration results to exclude parameter disturbances caused by atypical cooling processes such as abnormal door opening and closing, strong airflow, or short-term external heat sources approaching the panel.
[0053] When both the time constant deviation and thermal resistance deviation fall within the update range, the processor generates the updated thermal time constant and updated equivalent thermal resistance coefficient using a weighted average method. If the locally stored thermal time constant is... The fitted thermal time constant is The weighting coefficient is Then update the thermal time constant. Press We obtain; if the equivalent thermal resistance coefficient stored locally is The fitted equivalent thermal resistance coefficient is Then update the equivalent thermal resistance coefficient. Press Obtain. Weighting coefficients. The data is dimensionless and can range from 0.6 to 0.9. A larger value indicates that more of the original parameters are retained, while a smaller value indicates a faster response to the current cooling process. In the above formulas, both the thermal time constant and the fitted thermal time constant are in seconds, therefore the updated thermal time constant remains in seconds; both the equivalent thermal resistance coefficient and the fitted equivalent thermal resistance coefficient are in °C / W, therefore the updated equivalent thermal resistance coefficient remains in °C / W, and there is no mixing of dimensions. Finally, the processor writes the updated thermal time constant, the updated equivalent thermal resistance coefficient, and the average conduction power data into the local parameter table. The local parameter table can store multiple sets of data according to the target load, power segmentation, and update time, so that after the next target load is turned on, the dynamic compensation loop can read the equivalent thermal resistance coefficient and thermal time constant that are closer to the current environment and current load characteristics.
[0054] In the complete business process described above, the state transition moment provides the starting point for the dynamic compensation cycle, and the initial reference temperature provides the temperature reference at the moment of conduction. The real-time load current sequence and the real-time load voltage sequence form real-time active power data. After smoothing and power segmentation, the real-time active power data participates in the theoretical self-heating temperature rise recursion. The theoretical self-heating temperature rise is subtracted from the real-time onboard temperature at the same sampling time to form the decoupled ambient temperature. The decoupled ambient temperature is then compared with the target temperature range to form the adjustment command. After the target load is disconnected, the cooling temperature sequence is used to extract temperature decay characteristics and finally update the equivalent thermal resistance coefficient and thermal time constant. The entire data chain, from on / off identification, temperature decoupling, adjustment recording to parameter self-learning, retains the advantages of convenient onboard temperature data acquisition while reducing the impact of target load self-heating on the environmental perception results.
[0055] Furthermore, this application organically combines technical features such as identifying the on / off state of the target load, extracting the initial reference temperature at the moment of conduction, synchronously acquiring the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence in the dynamic compensation cycle, and recursively calculating the theoretical self-heating temperature rise based on real-time active power data. This enables the intelligent switch panel to establish a continuous correlation between load operating state, power change, panel self-heating, and ambient temperature around the same time axis. Among these features, the onboard temperature data acquired at the moment of conduction provides an initial reference for subsequent temperature judgment without significant self-heating interference; the real-time load current sequence and real-time load voltage sequence are used to reflect the actual power input of the target load at each sampling moment; the equivalent thermal resistance coefficient and thermal time constant convert this power input into a theoretical self-heating temperature rise with thermal inertia characteristics; based on this, the real-time onboard temperature is subtracted from the theoretical self-heating temperature rise to obtain the decoupled ambient temperature, so that the adjustment command is no longer directly triggered by the onboard temperature affected by mixed self-heating. Furthermore, after the target load is disconnected, a thermal environment self-learning cycle is entered. The equivalent thermal resistance coefficient and thermal time constant are updated in reverse using the temperature decay characteristics in the cooling temperature sequence, allowing the heat dissipation information obtained in the previous operation to be fed back to the next dynamic compensation cycle. Thus, the various technical features are not executed in isolation, but rather form a closed-loop collaborative mechanism from establishing the conduction reference, self-heating deduction during operation, decoupling from ambient temperature, generating adjustment commands, to parameter self-learning after shutdown. This improves the accuracy of current adjustment judgments and continuously enhances the adaptability of subsequent compensation calculations, thereby reducing the interference of load power fluctuations, installation environment differences, and long-term changes in operating conditions on the environmental perception results.
[0056] The working process and principle of this application are as follows: During operation, the intelligent switch panel continuously monitors the target load status connected to the panel. When the target load is detected to switch from an open to a closed state, the onboard temperature data collected at the moment of connection is extracted as the initial reference temperature to obtain a starting reference that is as close as possible to the actual ambient temperature. Subsequently, a dynamic compensation loop is entered, and the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence are acquired synchronously at each sampling moment. The real-time load current and real-time load voltage at the same sampling moment are multiplied to obtain the real-time active power data. Then, combined with the locally stored equivalent thermal resistance coefficient and thermal time constant, the theoretical self-heating temperature rise value is recursively calculated according to the heat accumulation and heat dissipation attenuation caused by the power change at adjacent sampling moments. Since this theoretical self-heating temperature rise value characterizes the additional temperature rise of the internal components and circuits of the panel due to power input after the target load is turned on, Therefore, by subtracting the theoretical self-heating temperature rise from the real-time onboard temperature at the corresponding sampling moment, the decoupled ambient temperature after eliminating the influence of panel self-heating can be obtained. The smart switch panel then compares the decoupled ambient temperature with the target temperature range and generates adjustment commands that match the comparison results, such as cooling, heating, or maintaining the temperature. When the target load is detected to switch from the on to the off state, the dynamic compensation cycle ends, and the panel enters the thermal environment self-learning cycle. During the natural cooling process after the target load stops heating, the cooling temperature sequence is collected, and the temperature decay characteristics are extracted to reflect the actual thermal response law under the current installation environment and heat dissipation conditions. This updates the equivalent thermal resistance coefficient and thermal time constant, making the calculation of the theoretical self-heating temperature rise more consistent with reality during the next on-run. This forms a closed-loop working mechanism of "on-run identification - self-heating compensation - environmental decoupling - adjustment output - cooling learning - parameter update". Example
[0057] This application also discloses an adaptive adjustment system for an intelligent switch panel based on environmental perception.
[0058] Reference Figure 2 An environmentally-aware intelligent switch panel adaptive adjustment system includes: The panel monitoring module is used to monitor the target load connected to the panel as it switches from disconnected to on state, and extracts the onboard temperature data collected at the moment of on-state as the initial reference temperature. The compensation loop module is used to enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The recursive calculation module is used to multiply the real-time load current and real-time load voltage at the same sampling time to generate real-time active power data, read the equivalent thermal resistance coefficient and thermal time constant stored locally, and recursively calculate the theoretical self-heating temperature rise value according to the change of real-time active power data at adjacent sampling times. The decoupling calculation module is used to subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupling ambient temperature. The instruction generation module is used to compare the decoupled ambient temperature with the target temperature range and generate adjustment instructions that match the comparison results. The self-learning loop module is used to monitor the target load switching from the on to the off state, exit the dynamic compensation loop and enter the thermal environment self-learning loop. The extraction and update module is used to obtain the cooling temperature sequence in the thermal environment self-learning loop, extract the temperature decay characteristics from the cooling temperature sequence, and update the equivalent thermal resistance coefficient and thermal time constant according to the temperature decay characteristics.
[0059] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention, they should all fall within the protection scope of the present invention.
[0060] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0061] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.
Claims
1. An adaptive adjustment method for an intelligent switch panel based on environmental perception, characterized in that, Includes the following steps: When the target load on the monitoring access panel switches from disconnected to on state, the onboard temperature data collected at the moment of on-state is extracted as the initial reference temperature. Enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The real-time load current and real-time load voltage at the same sampling time are multiplied to generate real-time active power data. The equivalent thermal resistance coefficient and thermal time constant stored locally are read, and the theoretical self-heating temperature rise is calculated recursively according to the change of real-time active power data at adjacent sampling times. Subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupled ambient temperature. The decoupled ambient temperature is compared with the target temperature range, and an adjustment command matching the comparison result is generated. When the target load is switched from on to off, the dynamic compensation loop is exited and the thermal environment self-learning loop is entered. The cooling temperature sequence is obtained in the thermal environment self-learning loop, the temperature decay characteristics are extracted from the cooling temperature sequence, and the equivalent thermal resistance coefficient and thermal time constant are updated according to the temperature decay characteristics. 2.The environment-aware based smart switch panel adaptive adjustment method of claim 1, wherein, When the target load on the monitoring access panel switches from disconnected to on, the onboard temperature data collected at the moment of on-state is extracted as the initial reference temperature, including: The real-time load current sequence of the target load is obtained according to a fixed sampling frequency; Divide the real-time load current sequence into adjacent pre-time windows and post-time windows, and calculate the mean pre-time current within the pre-time window and the mean post-time current within the post-time window respectively. The change in current is obtained by subtracting the average value of the preceding current from the average value of the subsequent current. When the change in current is greater than the preset jump threshold, the target load is determined to switch from the disconnected state to the on state, and the starting point of the subsequent time window is marked as the state switching time. Extract onboard temperature data within a preset time period before the state switching time and combine them into an environmental baseline sequence; The highest and lowest temperature data in the environmental reference sequence are removed, and the arithmetic mean of the remaining onboard temperature data in the environmental reference sequence is calculated to generate the initial reference temperature. 3.The environment-aware intelligent switch panel adaptive adjustment method of claim 2, wherein, Entering the dynamic compensation loop, real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence are synchronously acquired according to the sampling time, including: Using the state transition time as the starting point of the loop, a sampling time sequence is generated according to the compensation sampling interval; The onboard temperature data, load current data, and load voltage data collected at each sampling moment are bound into a single frame of operating condition data; The single-frame operating condition data is sorted according to the sampling time sequence to obtain the operating condition data sequence; In the operating condition data sequence, abnormal frames with time intervals greater than the compensation sampling interval are identified, and the onboard temperature data, load current data and load voltage data corresponding to the abnormal frames are supplemented by the average of the same type of data in adjacent valid frames. The real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence were extracted from the completed operating condition data sequence. 4.The environment-aware smart switch panel adaptive adjustment method of claim 1, wherein, Real-time active power data is generated by multiplying the real-time load current and real-time load voltage at the same sampling time, including: The real-time load current and real-time load voltage at the same sampling time are multiplied to generate single-point power data. The single-point power data are arranged in chronological order of sampling time to form a real-time active power sequence; Calculate the power mean of the real-time active power sequence within a predetermined number of bits before the current sampling time to generate smoothed active power data; The smoothed active power data is matched with the locally stored power segmentation table to determine the power segment corresponding to the target load. Read the power correction coefficient corresponding to the power segment, and multiply the power correction coefficient by the equivalent thermal resistance coefficient to obtain the segment equivalent thermal resistance coefficient. 5.The environment-aware based smart switch panel adaptive adjustment method of claim 4, wherein, The theoretical self-heating temperature rise is calculated recursively based on the change in real-time active power data at adjacent sampling times, including: Extract the smoothed active power data at the current sampling time and the theoretical self-heating temperature rise value at the previous sampling time; Based on smoothed active power data, the time interval between adjacent sampling times, and the piecewise equivalent thermal resistance coefficient, the heat generation increment at the current sampling time is calculated. Based on the theoretical self-heating temperature rise at the previous sampling time, the time interval between adjacent sampling times, and the thermal time constant, the heat dissipation attenuation at the current sampling time is calculated. Add the heat generation increment to the theoretical self-heating temperature rise at the previous sampling time, and subtract the heat dissipation attenuation to obtain the theoretical self-heating temperature rise at the current sampling time. Write the theoretical self-heating temperature rise value at the current sampling time into the self-heating temperature rise sequence. 6.The environment-aware based smart switch panel adaptive adjustment method of claim 5, wherein, Subtracting the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence yields the decoupled ambient temperature, including: Extract the real-time onboard temperature at the current sampling moment from the real-time onboard temperature sequence; Extract the theoretical self-heating temperature rise value at the current sampling moment from the self-heating temperature rise sequence; The single-point decoupled ambient temperature is generated by subtracting the theoretical self-heating temperature rise at the current sampling time from the real-time onboard temperature at the current sampling time. The single-point decoupled ambient temperatures are arranged in chronological order of sampling time to form a decoupled ambient temperature sequence; Calculate the difference between adjacent single-point decoupled ambient temperatures in the decoupled ambient temperature sequence to obtain the ambient temperature difference sequence; When the positive and negative directions of the differences in the predetermined number of positions in the ambient temperature difference sequence are the same and the cumulative difference is greater than the preset temperature threshold, the cumulative difference is incorporated into the initial reference temperature to generate a corrected reference temperature. 7.The environment-aware based smart switch panel adaptive adjustment method of claim 6, wherein, The decoupled ambient temperature is compared with the target temperature range, and adjustment commands matching the comparison results are generated, including: Read the lower and upper temperature limits of the target temperature range; Subtract the upper temperature limit from the single-point decoupled ambient temperature at the current sampling time to obtain the upper limit deviation data; Subtract the single-point decoupled ambient temperature at the current sampling time from the lower limit value to obtain the lower limit deviation data; When the upper limit deviation data is greater than zero, a cooling adjustment command is generated; When the lower limit deviation data is greater than zero, a temperature-increasing adjustment command is generated; When both the upper limit deviation data and the lower limit deviation data are not greater than zero, a hold-type adjustment command is generated; The adjustment command is combined with the current sampling time, the single-point decoupled ambient temperature, and the corrected reference temperature to form an adjustment record. 8.The environment-aware smart switch panel adaptive adjustment method of claim 1, wherein, When the monitored target load switches from an on to an off state, the dynamic compensation loop exits and enters a thermal environment self-learning loop, including: Obtain the real-time load current sequence of the target load in the dynamic compensation loop; Divide the real-time load current sequence into adjacent conduction end windows and disconnection confirmation windows, and calculate the average end current in the conduction end window and the average confirmation current in the disconnection confirmation window respectively. The change in turn-off current is obtained by subtracting the average value of the confirmed current from the average value of the end current. When the change in turn-off current is greater than the preset turn-off threshold, the target load is determined to switch from the on state to the off state. Mark the start point of the disconnect confirmation window as the disconnection time; Read the real-time onboard temperature, theoretical self-heating temperature rise, and decoupled ambient temperature at the time of shutdown, and generate an initial cooling record. The thermal environment self-learning loop is initiated based on the initial cooling record. 9.The environment-aware based smart switch panel adaptive adjustment method of claim 8, wherein, The cooling temperature sequence is obtained in the thermal environment self-learning loop, and temperature decay features are extracted from the cooling temperature sequence, including: Using the shutdown moment as the starting point for cooling sampling, onboard temperature data is continuously collected according to the cooling sampling interval to form a cooling temperature sequence. Calculate the temperature difference between adjacent onboard temperature data in the cooling temperature sequence to obtain the cooling difference sequence; When the absolute value of the temperature difference of a predetermined number in the cooling difference sequence is less than the cooling stability threshold, the average value of the corresponding onboard temperature data is calculated to generate the cooling endpoint temperature. Subtract the cooling endpoint temperature from each onboard temperature data in the cooling temperature sequence to obtain the cooling residual temperature sequence; Divide each cooling residual temperature data in the cooling residual temperature sequence by the theoretical self-heating temperature rise value corresponding to the turn-off time to obtain the normalized decay sequence. The normalized decay sequence is fitted with exponential decay to extract temperature decay features. 10.The environment-aware intelligent switch panel adaptive adjustment method of claim 9, wherein, The equivalent thermal resistivity and thermal time constant are updated based on temperature decay characteristics, including: The fitted thermal time constant is determined based on the temperature decay characteristics; Read the real-time load current sequence and real-time load voltage sequence arranged according to the sampling time during this conduction period; Multiply the real-time load current and real-time load voltage at the same sampling time to generate a conduction power sequence; Calculate the arithmetic mean of the conduction power sequence to generate the average conduction power data; Divide the theoretical self-heating temperature rise corresponding to the turn-off time by the average conduction power data to generate the fitted equivalent thermal resistance coefficient. Calculate the time constant deviation between the fitted thermal time constant and the locally stored thermal time constant, and calculate the thermal resistance deviation between the fitted equivalent thermal resistance coefficient and the locally stored equivalent thermal resistance coefficient. When both the time constant deviation and the thermal resistance deviation fall within the update range, the updated thermal time constant and the updated equivalent thermal resistance coefficient are generated by weighted averaging. Update the thermal time constant, update the equivalent thermal resistance coefficient, and write the average conduction power data to the local parameter table.
11. The intelligent switch panel adaptive adjustment system based on environmental perception, applied to the intelligent switch panel adaptive adjustment method based on environmental perception in any one of claims 1-10, characterized in that, include: The panel monitoring module is used to monitor the target load connected to the panel as it switches from disconnected to on state, and extracts the onboard temperature data collected at the moment of on-state as the initial reference temperature. The compensation loop module is used to enter the dynamic compensation loop and synchronously acquire the real-time onboard temperature sequence, real-time load current sequence, and real-time load voltage sequence according to the sampling time. The recursive calculation module is used to multiply the real-time load current and real-time load voltage at the same sampling time to generate real-time active power data, read the equivalent thermal resistance coefficient and thermal time constant stored locally, and recursively calculate the theoretical self-heating temperature rise value according to the change of real-time active power data at adjacent sampling times. The decoupling calculation module is used to subtract the theoretical self-heating temperature rise from the real-time onboard temperature corresponding to the sampling time in the real-time onboard temperature sequence to obtain the decoupling ambient temperature. The instruction generation module is used to compare the decoupled ambient temperature with the target temperature range and generate adjustment instructions that match the comparison results. The self-learning loop module is used to monitor the target load switching from the on to the off state, exit the dynamic compensation loop and enter the thermal environment self-learning loop. The extraction and update module is used to obtain the cooling temperature sequence in the thermal environment self-learning loop, extract the temperature decay characteristics from the cooling temperature sequence, and update the equivalent thermal resistance coefficient and thermal time constant according to the temperature decay characteristics.
