Defrosting control method for air source heat pump

Abstract

The invention relates to the technical field of air source heat pump control, and discloses an air source heat pump defrosting control method. According to the method, a frosting risk index and a trigger threshold are dynamically calculated through multi-sensor real-time data fusion processing; the defrosting effect is evaluated based on the defrosting process monitoring data, and defrosting duration parameters are adjusted in a self-adaptive mode; and finally, accurate triggering and termination of the defrosting control process are achieved by optimizing termination threshold judgment and system closed-loop feedback updating. The defrosting control accuracy and the system operation efficiency are effectively improved, the energy consumption is reduced, and the operation stability and reliability of the heat pump system are guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of air source heat pump control, and particularly relates to a defrosting control method for air source heat pump. BACKGROUND

[0002] In the technical field of air source heat pump control, the existing solutions related to defrosting control method generally perform defrosting operation based on single sensor threshold determination or fixed cycle, which has limitations such as defrosting trigger time deviation, parameter adjustment lag and insufficient dynamic response. The existing methods mostly rely on single indicators such as evaporator temperature or system pressure for stage determination, which is prone to defrosting demand misjudgment and execution time length misalignment in the scene of environmental humidity fluctuation and multi-working condition switching, and is difficult to meet the demand of heat pump system energy efficiency balance and heating stability. For the joint processing of evaporator temperature, environmental humidity and system pressure data, the existing technology generally lacks real-time fusion analysis capability of multi-source data trend, and is difficult to form a continuous control link of data acquisition-index calculation-effect evaluation-threshold optimization in the defrosting process dynamic adjustment, resulting in low matching degree of defrosting time length and actual demand and increased system energy consumption. SUMMARY

[0003] The present application provides a defrosting control method for air source heat pump to solve the problem of how to accurately determine the defrosting demand in the defrosting control process based on evaporator temperature, environmental humidity and system pressure data through multi-sensor data fusion technology.

[0004] In order to solve the above technical problems, the present application provides a defrosting control method for air source heat pump, comprising: Real-time data is obtained from the evaporator temperature sensor, the environmental humidity sensor and the system pressure sensor, time stamp alignment is performed based on the synchronous clock, and abnormal value elimination and unit standardization processing are performed based on threshold range detection, mutation detection and continuity detection, to generate fusion input data; Based on the fusion input data, temperature, humidity and pressure characteristic values are extracted, and a weighted algorithm calculation is performed based on normalization processing and dynamic adjustment weight parameters, to generate a frosting risk index and a dynamic trigger threshold, and a threshold mapping table is constructed based on time sequence form and incremental update mechanism; The dynamic trigger threshold is obtained from the threshold mapping table, the frosting risk index is compared based on point-by-point comparison and counter mechanism, and the duration is determined based on sliding time window algorithm, to generate a control signal; The control signal is obtained, the change rate is calculated based on numerical differentiation method and weighted filtering algorithm, and the defrosting effect is evaluated based on sliding window technology and multi-factor fusion algorithm, to generate a correction coefficient and adjust the preset defrosting time length parameter based on multiplication or addition model, to generate an optimized termination threshold; Based on the correction coefficient, the duration verification is performed based on the high-resolution timer and the sliding window algorithm in combination with the optimized termination threshold, and the state switching is performed based on the state management module and the preset control strategy to generate a termination instruction; The termination instruction is received, system operation parameters are obtained, the heat pump mode switching is performed based on the communication interface module and the preset control logic, the feedback coefficient calculation is performed based on the configuration analysis module and the feedback algorithm model, the restart configuration and the threshold adjustment suggestion are generated, and the system state closed-loop update is completed based on the parameter verification module and the multi-version management.

[0005] Further, the process of generating the fusion input data further includes: Real-time data of the evaporator temperature sensor, the environmental humidity sensor, and the system pressure sensor are obtained, time stamp alignment processing is performed based on the synchronous clock, and original sensor data are obtained; Temperature values, humidity values, and pressure values are extracted from the original sensor data, abnormal value elimination and unit standardization are performed based on threshold range detection, mutation detection, and continuity detection, and cleaned data are generated; The cleaned data are subjected to multi-source data synchronization processing based on the time window sliding algorithm, and fusion input data are generated.

[0006] Further, the process of constructing the threshold mapping table further includes: Temperature, humidity, and pressure characteristic values are extracted from the fusion input data, weighted algorithm calculation is performed based on normalization processing and dynamic adjustment of weight parameters, and a frost risk index is generated; Based on the frost risk index, dynamic threshold parameter calculation is performed based on historical frost risk index data and environmental parameters, and a dynamic trigger threshold is generated; The frost risk index and the dynamic trigger threshold are associatedly stored based on the time series form and the incremental update mechanism, and a threshold mapping table is generated.

[0007] Further, the process of generating the control signal further includes: The current dynamic trigger threshold is extracted from the threshold mapping table, the frost risk index comparison is performed based on the point-by-point comparison and the counter mechanism, and a threshold exceeding state flag is generated; Based on the sliding time window algorithm and the priority rule, the defrosting necessity determination is performed according to the duration of the threshold exceeding flag, and a start instruction is generated; The start instruction is subjected to execution priority sorting based on the multi-level priority queue management mechanism, and a control signal is generated.

[0008] Further, the process of generating the optimized termination threshold further includes: The control signal is obtained, the change rate calculation is performed based on the numerical differentiation method and the weighted filtering algorithm, and a process indicator is generated; Extract the temperature recovery rate and pressure stability value from the process index, evaluate the defrosting effect based on the sliding window technology and multi-factor fusion algorithm, and generate a correction coefficient; Adjust the preset defrosting duration parameter based on the multiplication or addition model according to the correction coefficient, and generate an optimized termination threshold.

[0009] Further, the process of generating the termination instruction further includes: Obtain the correction coefficient, combine the optimized termination threshold, and generate the termination candidate marker based on the timestamp alignment algorithm and sliding time window technology; Verify the duration of the termination candidate marker based on the high-resolution timer and sliding window algorithm, and generate a compliance signal; According to the compliance signal, trigger the system state switching based on the state management module and the preset control strategy, and generate the termination instruction.

[0010] Further, the process of completing the system state closed-loop update based on the parameter verification module and multi-version management further includes: After receiving the termination instruction, obtain the current system running parameter, perform heat pump mode switching processing based on the communication interface module and the preset control logic, and generate a restart configuration; Extract the key running parameters from the restart configuration, calculate the feedback coefficient based on the configuration parsing module and the feedback algorithm model, and generate a threshold adjustment suggestion; Write the threshold adjustment suggestion into the dynamic threshold adjustment algorithm parameter library based on the parameter verification module and multi-version management, and complete the system state closed-loop update.

[0011] Further, the expression of generating the optimized termination threshold further includes: Calculate the temperature change rate by the improved cross-correlation algorithm : Wherein, is the temperature change rate; is the index number of the current data point; is the summation index variable; is the sliding window length; is the evaporator surface temperature; is the timestamp; is the time delay parameter; is the temperature mean value in the window; is the temperature difference between the beginning and end of the window; is the time interval; Calculate the pressure change rate by pressure difference: Wherein, is the pressure change rate; is the refrigerant pressure; is a differential order; Further, the defrosting effect is evaluated by dynamic weighting fusion: wherein, is a defrosting effect evaluation value; is a weight coefficient; is a temperature rise rate historical extreme value; is a pressure stability value; is a pressure stability historical extreme value; The correction coefficient is generated by an S-shaped function: wherein, is a correction coefficient; is a slope parameter; is an effect reference value; Based on the correction coefficient , the preset defrosting duration is adjusted : wherein, is an adjusted defrosting duration; is a preset defrosting duration; is an adjustment amplitude coefficient; An optimization termination threshold is constructed : wherein, is an optimization termination threshold; is a preset threshold; is a sensitivity parameter.

[0012] Further, the process of adjusting the preset defrosting duration parameter further comprises: Combined with the correction coefficient, the preset duration is adjusted by a multiplication or addition model, and the specific adjustment mode is segmented according to the numerical range of the correction coefficient.

[0013] Further, the process of generating the optimization termination threshold further comprises: According to the defrosting duration parameter, the optimized termination determination threshold is calculated, and the calculation of the termination determination threshold comprehensively considers the temperature rise rate, the pressure stability and the environmental humidity change, forming a multi-dimensional determination standard.

[0014] The key innovations of the present application include: (1) Through the multi-sensor data fusion technology, combined with the evaporator temperature, environmental humidity and system pressure data, a dynamic acquisition and real-time analysis defrosting control link is established to accurately determine the defrosting demand.

[0015] (2) The defrosting effect is evaluated by using the temperature recovery rate and the pressure stable value in the process index, a correction coefficient is generated, and dynamic adjustment and parameter optimization of the defrosting process are realized.

[0016] (3) The preset defrosting time parameter is adjusted according to the correction coefficient, an optimized termination threshold is generated, and a self-adaptive control link of the defrosting process is formed, so that the matching degree of the defrosting time and the actual demand is improved.

[0017] The main beneficial effects are as follows: (1) Through the multi-sensor data fusion technology, the defrosting control link can realize dynamic acquisition and real-time analysis of data in the defrosting demand judgment process, effectively avoid defrosting trigger time deviation and misjudgment, and is suitable for environmental humidity fluctuation and multi-working condition switching scenes.

[0018] (2) The generation and application of the correction coefficient enable the defrosting process evaluation link to be dynamically adjusted based on the process index, realize real-time optimization of the defrosting effect, and be suitable for multi-stage adjustment requirements of the defrosting process.

[0019] (3) Through the generation and application of the optimized termination threshold, the self-adaptive control link can improve the matching degree of the defrosting time and the actual demand, reduce system energy consumption, and be suitable for heat pump system energy efficiency balance and heating stability requirements. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 A flowchart of a defrosting control method of an air source heat pump provided by the embodiment of the application. DETAILED DESCRIPTION

[0021] Embodiment one: refer to Figure 1 , is a flowchart of a defrosting control method of an air source heat pump provided by the embodiment of the application, which can at least include steps S100-S600: S100, real-time data is acquired from an evaporator temperature sensor, an environmental humidity sensor and a system pressure sensor, time stamp alignment is performed based on a synchronous clock, and abnormal value elimination and unit standardization processing are performed based on threshold range detection, mutation detection and continuity detection, to generate fusion input data; S200, based on the fusion input data, temperature, humidity and pressure characteristic value extraction and weighted algorithm calculation based on normalization processing and dynamic adjustment weight parameter are performed, a frosting risk index and a dynamic trigger threshold are generated, and a threshold mapping table is constructed based on a time sequence form and an incremental update mechanism; S300, the dynamic trigger threshold is acquired from the threshold mapping table, frosting risk index comparison and duration determination based on a sliding time window algorithm are performed based on point-by-point comparison and a counter mechanism, and a control signal is generated; S400, acquire the control signal, calculate the change rate based on the numerical differentiation method and the weighted filtering algorithm, evaluate the defrosting effect based on the sliding window technology and the multi-factor fusion algorithm, generate the correction coefficient, adjust the preset defrosting time parameter based on the multiplication or addition model, and generate the optimization termination threshold; S500, based on the correction coefficient, combine the optimization termination threshold, verify the duration based on the high-resolution timer and the sliding window algorithm, switch the state based on the state management module and the preset control strategy, and generate the termination instruction; S600, receive the termination instruction, acquire the system running parameter, switch the heat pump mode based on the communication interface module and the preset control logic, calculate the feedback coefficient based on the configuration analysis module and the feedback algorithm model, generate the restart configuration and the threshold adjustment suggestion, and complete the system state closed-loop update based on the parameter verification module and the multi-version management.

[0022] Step S100 includes at least steps S110-S130: S110, acquire the real-time data of the evaporator temperature sensor, the environment humidity sensor and the system pressure sensor, perform time stamp alignment processing based on the synchronous clock, and obtain the original sensor data; Real-time data of the evaporator temperature sensor, the ambient humidity sensor and the system pressure sensor are acquired as the input source of the present step. Specifically, the real-time data are continuously acquired by the acquisition module of each sensor, and each piece of data is attached with time stamp information to ensure the time sequence integrity of the data. The evaporator surface temperature measurement value in Celsius (℃) acquired by the evaporator temperature sensor is acquired; the ambient air humidity measurement value in relative humidity percentage (%RH) acquired by the ambient humidity sensor is acquired; and the refrigerant system pressure measurement value in kilopascal (kPa) acquired by the system pressure sensor is acquired. All real-time data are continuously acquired by the acquisition module built-in or dedicated to each sensor through a dedicated bus or wireless transmission mode, and each data packet is attached with high-precision time stamp information, which is derived from the system synchronous clock, to ensure the acquisition time sequence integrity of multi-source data in the time dimension and provide a structured raw data basis for subsequent data alignment, cleaning and fusion processing. The evaporator temperature sensor is arranged on the evaporator surface and can reflect the heat exchange state of the evaporator in real time; the ambient humidity sensor is installed in the ambient environment of the equipment to monitor the air humidity change; and the system pressure sensor is used to acquire the pressure state in the refrigerant system. During the data acquisition process, the system performs time stamp alignment processing on the sensor data through the synchronous clock to ensure the consistency of the data acquired by different sensors in the time dimension. To achieve efficient acquisition, the sensor data are transmitted to the data preprocessing unit through a dedicated bus or wireless transmission mode, and the data preprocessing unit monitors the data reception state. If data loss or abnormal interruption is found, an abnormal record mechanism is triggered to store abnormal event information for subsequent analysis. Specifically, the data acquisition frequency is dynamically adjusted according to the system setting parameters, which meets the real-time requirement and avoids data redundancy. The raw sensor data including the temperature value, the humidity value, the pressure value and the corresponding time stamp information are the output field name of the present step, which is called by the “data cleaning” of the subsequent step S120.

[0023] In S120, the temperature value, the humidity value and the pressure value are extracted from the raw sensor data, the abnormal values are removed based on threshold range detection, mutation detection and continuity detection, and the unit standardization is performed to generate the cleaned data. Further, temperature values, humidity values and pressure values are extracted from the original sensor data, and outlier rejection and unit standardization processing are performed. Specifically, the outlier rejection adopts multiple judgment mechanisms, including but not limited to threshold range detection, mutation detection and continuity detection. The threshold range detection filters measurement values outside the reasonable range based on historical data statistical characteristics and device calibration parameters; the mutation detection rejects instantaneous abnormal fluctuation data by comparing the change amplitude of adjacent sampling points; the continuity detection judges data loss or long-term static state. After outlier rejection, the data is subjected to unit standardization, which uniformly converts the physical quantities output by different sensors into standard units agreed by the system internally, so as to facilitate subsequent data fusion processing. In the standardization process, temperature is uniformly converted into Celsius (℃), humidity is converted into relative humidity percentage (%RH), and pressure is converted into kilopascal (kPa). The data cleaning process also includes data integrity verification, and interpolation algorithm is used for reasonable compensation for missing data to ensure data continuity. After cleaning, the generated cleaned data includes temperature, humidity and pressure feature values with outliers rejected and units unified, and is attached with corresponding time stamp as the output field name of this step, which is called by the subsequent S130 "data fusion".

[0024] S130, multi-source data synchronization processing is performed on the cleaned data based on a time window sliding algorithm to generate fusion input data. For the cleaned data, multi-source data synchronization processing is performed, specifically, temperature, humidity and pressure data are aligned in time dimension according to time stamp to form fusion input data. The time dimension alignment adopts a time window sliding algorithm, and a reasonable time window width is set to balance the data synchronization accuracy and processing efficiency. For multiple data in the time window, weighted average or latest value replacement strategy is adopted to ensure the representativeness and real-time performance of the fusion input data. Further, the fusion input data constructs a multi-dimensional data structure containing temperature, humidity and pressure three-dimensional features, which facilitates the input of subsequent weighted algorithm. During the data fusion processing, the system records the data synchronization state and abnormal situation to ensure the integrity and accuracy of the data link. The fusion input data is transmitted to the subsequent step S210 "weighted calculation" as the output field name of this step, to support the calculation of the frost risk index and the dynamic threshold adjustment. The generation of the fusion input data runs through the data collection, cleaning and synchronization processing links, realizes the unified management and high-quality input of multi-sensor data, and promotes the accurate implementation of the subsequent defrosting control strategy.

[0025] Step S200 includes at least steps S210-S230: S210, temperature, humidity and pressure feature values are extracted from the fusion input data, and weighted algorithm calculation is performed based on normalization processing and dynamic adjustment of weight parameters to generate a frost risk index; Based on the fusion input data, specifically, the fusion input data is generated by step S130 "data fusion", containing temperature, humidity and pressure characteristic values after time dimension alignment, as the input source of the step S210 "weighted calculation". The temperature characteristic value reflects the heat exchange state of the evaporator surface, the humidity characteristic value represents the environmental air humidity level, and the pressure characteristic value embodies the pressure state in the refrigerant system, all of which are attached with timestamp information to ensure the time sequence integrity of the data. Specifically, the weighted algorithm first normalizes the characteristic values to eliminate the influence of different physical quantity dimension differences on the calculation, and then sums the normalized temperature, humidity and pressure characteristic values according to the preset weight, the weight parameter is dynamically adjusted according to the historical operation data and environmental conditions, specifically including temperature weight, humidity weight and pressure weight, corresponding to the direct influence, auxiliary influence and system state influence of the frost risk respectively. In the process of weighted calculation, the system filters the abnormal fluctuations of the input data, adopts the combination of sliding window average and threshold truncation to eliminate the sudden abnormal data, and ensures the stability and continuity of the calculation result. Further, the weighted calculation also combines the time series data of the historical frost risk index, adopts the weighted moving average algorithm to smooth the fluctuation trend of the current frost risk index, and improves the reliability of the index. After completing the weighted calculation, the frost risk index is generated, which is a quantitative index that comprehensively reflects the possibility of frost on the evaporator temperature, environmental humidity and system pressure. The frost risk index is stored in the time series database to support the subsequent trend analysis and rapid access of the decision module. The frost risk index as the output field name of this step is transmitted to step S220 "threshold adjustment" for dynamic threshold parameter calculation.

[0026] S220, based on the frost risk index, based on the historical frost risk index data and environmental parameters, the dynamic threshold parameter calculation is generated, and the dynamic trigger threshold is generated. Specifically, step S220 "threshold adjustment" takes the frost risk index generated by step S210 "weighted calculation" as input, and combines the environmental temperature change rate and humidity saturation as auxiliary input parameters. The environmental temperature change rate is obtained by continuous time difference calculation of the environmental temperature sensor data, reflecting the change trend of the environmental temperature, and the humidity saturation is calculated based on the physical relationship between relative humidity and environmental temperature, reflecting the saturation degree of water vapor content in the air. Specifically, the dynamic threshold parameter calculation adopts an adaptive algorithm model, which is based on multivariate regression analysis or machine learning method, combines historical frost risk index data and environmental parameters, and adjusts the defrosting trigger threshold in real time. The calculation process first extracts the features of the input environmental temperature change rate and humidity saturation, including mean, variance and change trend, and then inputs the features into the adaptive threshold model to output the corresponding dynamic trigger threshold. The model is internally constrained by upper and lower limits of the threshold to prevent the threshold from deviating from the reasonable range. During the dynamic threshold parameter calculation process, the system continuously monitors the integrity and effectiveness of the input data, and starts the abnormal record mechanism for abnormal data to ensure the accuracy of the model input. Further, the dynamic trigger threshold is combined with the current seasonal characteristics and the equipment operating state for weighted adjustment, reflecting the influence of environmental changes on the defrosting demand. After calculation, the adaptive trigger threshold is generated as the output field name of this step, which is passed to step S230 "threshold update" for associated storage of the threshold and the frost risk index.

[0027] S230, the frost risk index and the dynamic trigger threshold are associated and stored based on time series form and incremental update mechanism to generate a threshold mapping table; In step S230 "threshold updating", specifically, the frost risk index generated in step S210 "weighted calculation" is associated with the dynamic trigger threshold generated in step S220 "threshold adjustment" for storage, forming a threshold-index mapping table. The mapping table is organized in time series form, containing timestamp, frost risk index value, dynamic trigger threshold value and state marker field, supporting multi-condition query and historical data backtracking. Specifically, the mapping table adopts an incremental updating mechanism. When adding new data items, the system first checks the continuity and uniqueness of the timestamp to prevent data conflicts, and then writes the latest frost risk index and corresponding dynamic trigger threshold to the mapping table. The mapping table structure design supports fast retrieval, facilitating real-time access by the subsequent defrosting determination module. Further, the system periodically checks and cleans the data in the mapping table, eliminating outdated or abnormal data to ensure data integrity and accuracy. The state marker field is used to identify the exceeding state of the current frost risk index relative to the dynamic trigger threshold, assisting in the determination of the defrosting trigger condition. The updating process of the mapping table includes anomaly detection and log recording to ensure the stable operation of the data link. The threshold mapping table is the output field name of this step, which is passed to step S310 "threshold comparison" for determination of the defrosting trigger condition. Through the above steps, the association management of the frost risk index and the dynamic threshold is completed, forming the core data basis of the multi-source data fusion and threshold dynamic adjustment module.

[0028] Step S300 includes at least steps S310-S330: S310, extract the current dynamic trigger threshold from the threshold mapping table, compare the frost risk index based on point-by-point comparison and counter mechanism, and generate a threshold exceeding state marker; The threshold mapping table output from step S230 "threshold update" is taken as input, and step S310 "threshold comparison" specifically extracts the dynamic triggering threshold corresponding to the current time and its associated frost risk index from the threshold mapping table. The threshold mapping table contains a timestamp, a frost risk index value, a dynamic triggering threshold value, and a state marker field. Step S310 first searches the mapping table to locate the latest valid data entry, ensuring that the extracted dynamic triggering threshold and frost risk index strictly correspond in time. Specifically, the comparison of the dynamic triggering threshold and the frost risk index uses a real-time reading mechanism combined with timestamp synchronization to avoid misjudgment due to data delay. The comparison process compares the frost risk index and the dynamic triggering threshold point by point to determine whether the current frost risk index exceeds the dynamic triggering threshold. To ensure the robustness of the judgment, the system introduces a threshold exceeding state marker field that accumulates the results of multiple consecutive comparisons to avoid false triggering caused by single-point abnormal data. Specifically, the threshold exceeding state marker is implemented through a counter mechanism. When the frost risk index exceeds the dynamic triggering threshold for a preset time window length, the state marker switches from a non-exceeding state to an exceeding state. This mechanism reduces the impact of transient fluctuations on the control logic by statistically analyzing the state within the time window. During the comparison process, the system checks the integrity of the input data. If it finds that the mapping table data is missing or abnormal, it triggers an exception recording mechanism to record the abnormal event and uses the latest valid data for compensation processing. The threshold exceeding state marker is the output field name of this step, which is passed to step S320 "signal generation" for subsequent defrosting start determination. The processing of step S310 ensures real-time and accurate comparison of the dynamic threshold and the frost risk index, providing a key state identifier for defrosting trigger condition determination. The output of this step closely links with the input of step S320, supporting continuous closed-loop operation of defrosting control.

[0029] S320, according to the duration of the threshold exceeding marker, determine the necessity of defrosting based on the sliding time window algorithm and priority rules, and generate a start instruction; Based on the threshold exceeding state flag generated in step S310 "threshold comparison", step S320 "signal generation" specifically monitors and analyzes the duration of the state flag as the core basis for judging the necessity of defrosting. The threshold exceeding state flag records the start and end time of its state change through time stamp, and step S320 collects the time sequence data of the state flag and calculates the duration of its continuous exceeding state. Specifically, the system sets a predefined duration threshold as the time threshold for defrosting start, to prevent frequent start of the defrosting program due to temporary threshold exceeding. The duration determination adopts a sliding time window algorithm to dynamically adjust the monitoring interval to adapt to different environmental change rates. Further, step S320 combines the priority rules in the defrosting control strategy to make multi-condition determination on the duration of the threshold exceeding state flag, including environmental temperature, system operation mode and historical defrosting frequency and other auxiliary parameters, to ensure the rationality of the defrosting start instruction. During the determination process, the system filters the abnormal fluctuations of the input state flag, adopts a combination of weighted average and threshold truncation to eliminate occasional abnormalities, and improves the stability of the determination. If the duration of the threshold exceeding state flag meets the preset conditions, step S320 generates a defrosting start instruction, which includes a start time stamp, a priority level and control parameters, and is specifically used to trigger the execution of the defrosting program. The start instruction is the output field name of this step, which is transmitted to step S330 "instruction transmission" for subsequent generation and scheduling of control signals. The processing flow of step S320 ensures the timing accuracy and logical rigor of the defrosting start signal, prevents false triggering, improves the rationality of system response, and forms an effective transmission link with step S330, ensuring the continuity and coordination of defrosting control.

[0030] S330, performing priority sorting on the start instruction based on a multi-level priority queue management mechanism to generate a control signal; Step S330 "instruction delivery" takes the defrosting start instruction output by step S320 "signal generation" as input, and specifically performs execution priority sorting and timestamp management on the start instruction to generate a timestamped control signal. The execution priority sorting is based on preset priority rules, combined with the current system running state, defrosting emergency degree and historical defrosting execution records, to dynamically adjust the execution order of the start instruction. Specifically, the system adopts a multi-level priority queue management mechanism to classify and store and dispatch start instructions of different priorities, ensuring that high-priority instructions are executed first. In terms of timestamp management, step S330 adds precise timestamp information to the start instruction to record the specific time when the instruction is generated, supporting subsequent execution tracking and abnormality analysis. Further, the generation of the control signal includes instruction format conversion, communication protocol packaging and security verification to ensure the integrity and effectiveness of the control signal during transmission. Step S330 also integrates an abnormality detection mechanism to monitor the transmission state of the control signal in real time, triggers a retransmission strategy for transmission failure or delay, and records abnormal logs for subsequent diagnosis. The generated control signal contains all the information and scheduling parameters of the start instruction, which is the output field name of this step, and is delivered to step S410 "process monitoring" for real-time monitoring of defrosting execution. The processing of step S330 realizes the standardized management and accurate scheduling of the defrosting start instruction, ensures the orderly execution of the defrosting program, and forms a continuous closed-loop control link with the subsequent steps.

[0031] Step S400 includes at least steps S410-S430: S410, obtain the control signal, calculate the change rate based on the numerical differentiation method and the weighted filtering algorithm to generate the process index; Specifically, the control signal contains a timestamp and execution instruction of defrosting start, and the system starts real-time monitoring of the defrosting process according to the control signal. The temperature sensor and the pressure sensor are arranged on the surface of the evaporator and the key node of the refrigerant system respectively, and continuously collect temperature values and pressure values during the defrosting execution, and the collected data are all attached with timestamp information to ensure the time sequence integrity of the data. Specifically, the system receives the sensor data through a high-speed data bus or a wireless communication protocol, and stores the data in a data buffer area in real time. In order to counteract the dynamic changes of the defrosting process, the system calculates the change rate of the temperature and pressure data, which specifically includes difference operation and sliding window average processing of continuous sampling points, and filters out noise and abnormal fluctuations. The change rate calculation adopts a numerical differentiation method based on time series, combined with a weighted filtering algorithm, to smooth the burst signal and ensure the stability and response sensitivity of the change rate index. Further, the change rate calculation considers the non-uniformity of the sampling time interval, adjusts the difference result through a time weighting factor, and improves the accuracy of the calculation. After the calculation is completed, the system constructs a defrosting progress index according to the combination characteristics of the temperature change rate and the pressure change rate, which comprehensively reflects the temperature rising trend of the evaporator surface and the stable state of the refrigerant pressure, and can dynamically depict the progress of the defrosting process. The defrosting progress index is stored in the form of time series data, which supports subsequent evaluation of the defrosting effect. The progress index is taken as the output field name of the present step, and is transmitted to step S420 "progress evaluation" for further analysis of the defrosting effect.

[0032] S420, extract the temperature rising rate and the pressure stable value from the progress index, evaluate the defrosting effect based on the sliding window technology and the multi-factor fusion algorithm, and generate a correction coefficient; Step S420 "extracting temperature recovery rate and pressure stability value from the process indicators, evaluating defrosting effect, and generating correction coefficient" takes step S410 "process indicators" as input. Specifically, the process indicators include time series data of temperature change rate and pressure change rate. The system first extracts features from the process indicators, including calculating the average value, maximum value, and change trend of the temperature recovery rate, which reflects the recovery speed of the evaporator surface temperature during the defrosting process and is directly related to the quality of the defrosting effect. The pressure stability value is obtained by calculating the variance and standard deviation of the pressure change rate, reflecting the fluctuation amplitude of the refrigerant system pressure, and the pressure stability is an important basis for judging whether the defrosting process is smooth. The sliding window technique is used for feature extraction, and the window width is dynamically adjusted according to the system operating parameters, taking into account real-time performance and data stability. Further, the system normalizes the extracted feature values to eliminate the dimensional influence between different physical quantities, facilitating subsequent comprehensive evaluation. Based on the normalized temperature recovery rate and pressure stability value, the system uses a multi-factor fusion algorithm to evaluate the defrosting effect. Specifically, the fusion algorithm combines weight coefficients to weight and sum each feature value. The weight coefficients are pre-set and can be dynamically adjusted according to historical operating data, reflecting the contribution of each indicator to the defrosting effect. During the evaluation process, the system detects the integrity and abnormal fluctuations of the input data. Abnormal data is marked and removed by an anomaly detection model to avoid misleading the evaluation results. The defrosting effect evaluation result is converted into a correction coefficient through a mapping function. The correction coefficient is used to adjust the defrosting duration parameter, and the range of the correction coefficient is set with upper and lower limit constraints to prevent excessive parameter adjustment. The correction coefficient is output in numerical form with timestamp information to support time sequence management in subsequent steps. The correction coefficient is the output field name of this step and is passed to step S430 "parameter adjustment" for optimization adjustment of the defrosting duration.

[0033] S430, adjusting the preset defrosting duration parameter based on a multiplication or addition model according to the correction coefficient to generate an optimized termination threshold; Step S430 "adjusting preset defrosting duration parameter according to the correction coefficient to generate optimized termination judgment threshold" takes step S420 "correction coefficient" as input. Specifically, the correction coefficient is used to dynamically adjust the preset defrosting duration parameter based on the real-time evaluation of defrosting effect. The system first reads the current preset defrosting duration parameter from the parameter library, which is set based on the device model and historical operation experience, serving as the basis for defrosting termination judgment. Combined with the correction coefficient, the system adjusts the preset duration through multiplication or addition model. The specific adjustment method is segmented according to the numerical range of the correction coefficient to ensure the smoothness and rationality of the adjustment. During the adjustment process, the system performs boundary detection on the correction coefficient to prevent the adjusted defrosting duration from exceeding the safe operation range of the device. Further, the system calculates the optimized termination judgment threshold based on the adjusted defrosting duration parameter, combined with the current environment and system state. This threshold is used to judge the recovery degree of the frosting risk index in the subsequent process. The calculation of the termination judgment threshold considers temperature recovery rate, pressure stability, and environmental humidity changes, forming a multi-dimensional judgment standard. The system stores the optimized termination judgment threshold in the time series database, supporting real-time calling and historical data comparison. To ensure data consistency and integrity, the system performs data verification and anomaly detection during the adjustment process, triggering log recording and alarm mechanism for abnormal adjustment operations. The optimized termination threshold is the output field name of this step, which is passed to step S510 "termination condition" for defrosting termination judgment. Steps S410 to S430 form a closed-loop control link for real-time defrosting process monitoring and parameter optimization, supporting dynamic adjustment of defrosting duration and stability maintenance of system operation.

[0034] In another embodiment, in S410, the input sources are the defrosting start instruction and real-time data of temperature and pressure sensors passed from step S330. The system starts high-speed data acquisition according to the instruction and continuously samples the evaporator surface temperature (℃) and refrigerant pressure (kPa) during defrosting execution, with each data attached with a timestamp (s). To construct the defrosting progress index, formula ① calculates the temperature change rate : Wherein: is the temperature recovery rate, i.e. the temperature change rate, representing the instantaneous change rate of temperature at the current time; is the index number of the current data point; is the summation index variable; is the sliding window length; is the evaporator surface temperature; is the timestamp; is the time delay parameter; is the temperature mean value in the window; is the temperature difference between the beginning and the end of the window; is the time interval.

[0035] The data source is the real-time data of the temperature sensor. Formula ② calculates the pressure change rate through pressure difference: wherein: is the pressure change rate, representing the instantaneous change rate of the pressure at the current time; is the refrigerant pressure, indicating the pressure value sampled at time , the data coming from the pressure sensor; is the difference order (taking values 3-5), the data source being the pressure sensor data.

[0036] The final generated process indicator is passed to the “progress evaluation” of S420 as the output field name “process indicator”.

[0037] Further, in S420, the process indicator of S410 is accepted, the temperature recovery rate and the pressure stability value (the variance of the pressure change rate) are extracted. Formula ③ uses dynamic weighted fusion to evaluate the defrosting effect: wherein: is the defrosting effect evaluation value, which is a normalized comprehensive score; is the weight coefficient; is the temperature recovery rate historical extreme value; is the pressure stability value, specifically the variance of the pressure change rate , representing the fluctuation of the pressure change rate, the smaller the value, the more stable the pressure; is the pressure stability historical extreme value, the data source being the historical operation database.

[0038] Formula (IV) generates a correction coefficient through an S-shaped function: Wherein: is the correction coefficient; is a slope parameter; is an effect reference value, and the data source is a system preset parameter.

[0039] The output field name "correction coefficient" is passed to "parameter adjustment" of S430.

[0040] Further, in S430, based on the correction coefficient of S420 , formula (V) adjusts the preset defrosting duration : Wherein: is the adjusted defrosting duration; is the preset defrosting duration; is an adjustment amplitude coefficient, and the data source is a device safety constraint library.

[0041] Formula (VI) constructs an optimization termination threshold : Wherein: is the optimization termination threshold; is a preset threshold; is a sensitivity parameter, and the data source is a dynamic threshold model.

[0042] The output field name "optimization termination threshold" is passed to "termination condition" of S510.

[0043] Technical effect of this paragraph: through dynamic adjustment of the defrosting duration parameter and the threshold optimization, the adaptive matching of the defrosting termination condition is realized, and the system response efficiency and the operation stability are improved.

[0044] Step 500 at least contains steps S510-S530: S510, acquire the correction coefficient, combine the optimization termination threshold, generate the termination candidate mark based on the timestamp alignment algorithm and the sliding time window technology; The step S510 "termination condition" takes the optimization termination threshold output by the step S430 "parameter adjustment" as input, in combination with the real-time frosting risk index generated by the step S210 "weighting calculation". Specifically, the optimization termination threshold includes multi-dimensional judgment parameters adjusted based on the correction coefficient, and the frosting risk index reflects the comprehensive risk level of frosting under the current environmental and system state. The system first continuously acquires the latest frosting risk index data through the data acquisition module, ensuring that the timestamp of the data is synchronized with the optimization termination threshold. Specifically, the frosting risk index and the optimization termination threshold are strictly matched in the time dimension, and the timestamp alignment algorithm is adopted to prevent judgment errors caused by data delay or loss. Subsequently, the system monitors the downward trend of the frosting risk index in real time, calculates the change rate and trend stability of the index using the sliding time window technology, and determines whether the frosting risk index presents a continuous downward trend. This process combines threshold parameters to determine whether the current index has returned to the preset safe range. To avoid misjudgment, the system introduces a multiple judgment mechanism, including time length detection of the index continuously below the optimization termination threshold and index fluctuation amplitude analysis, to ensure the stability and accuracy of the judgment. During the judgment process, the system checks the integrity of the input data, and starts an abnormal record mechanism for abnormal data to save abnormal events and data state for subsequent analysis. Specifically, when the frosting risk index continuously decreases and stabilizes below the optimization termination threshold, the system generates a termination candidate marker, which includes a timestamp and judgment basis information, for subsequent stability verification. The termination candidate marker is the output field name of this step, which is passed to the "stability verification" of step S520, and at the same time, the marker forms a data link with the system state update of step S600, supporting closed-loop control of defrosting termination.

[0045] S520, continuously verify the duration of the termination candidate marker based on a high-resolution timer and a sliding window algorithm, and generate a compliance signal; Step S520 "stability verification" takes the termination candidate marker generated by step S510 "termination condition" as input, specifically, the termination candidate marker contains the frosting risk index recovery trend and timestamp information. The system first accurately calculates the duration of the marker, uses a high-resolution timer to record the start and end time of the marker state, and ensures the accuracy of the time data. Further, the system sets a predefined duration threshold as the threshold for judging the stability of the defrosting termination, to prevent false judgments caused by temporary index fluctuations. The duration verification uses a sliding window algorithm to dynamically adjust the monitoring interval length, adapting to different environmental change rates and system response characteristics. Specifically, the system performs continuity analysis on the time series of the termination candidate marker, removes discontinuous or abnormal periods, and ensures that the judgment is based on continuous and stable index recovery state. During the verification process, the system dynamically adjusts the duration threshold based on historical defrosting data and environmental parameters, improving the flexibility and adaptability of the judgment. To ensure the accuracy of the data, the system performs anomaly detection on the input marker, triggers an anomaly recording mechanism when an anomaly is found, and saves the anomaly information and related context data. After the duration verification, if the duration of the termination candidate marker meets the preset threshold, the system generates a stability compliance signal, which contains the verification timestamp and state identifier, indicating that the defrosting termination condition has reached a stable state. The compliance signal is the output field name of this step, which is passed to the "instruction generation" call of step S530, and the signal forms a closed-loop feedback with the system state update module of step S600, supporting the safe termination of the defrosting process.

[0046] S530, according to the compliance signal, triggers system state switching based on the state management module and the preset control strategy, and generates a termination instruction; Step S530 "instruction generation" takes the stability compliance signal output by step S520 "stability verification" as input. Specifically, the compliance signal indicates that the frost risk index has been stably restored to the safe range, and the conditions for terminating defrosting are met. After receiving the signal, the system first performs final confirmation of the instruction triggering condition, combines the current system operation state, environmental parameters and equipment safety limits to ensure the rationality of the termination instruction. Specifically, the system reads the current defrosting execution state and historical instruction execution record through the state management module to prevent the generation of repeated or conflicting instructions. Subsequently, the system generates a defrosting termination instruction according to the pre-set control strategy, and the instruction content includes the termination timestamp, termination type identification and related control parameters, supporting flexible switching of multiple defrosting termination modes. The termination instruction is formatted to meet the system communication protocol and safety verification requirements, ensuring the integrity and effectiveness of the instruction during transmission. The system manages the priority of the termination instruction, combines the current system load and operation priority to reasonably schedule the sending time of the instruction. Further, the termination instruction generation process integrates an abnormality monitoring mechanism to monitor the instruction generation and transmission state in real time, triggers a retransmission or alarm mechanism for abnormal conditions, and records detailed logs for subsequent analysis. The generated defrosting termination instruction is the output field name of this step, which is delivered to step S610 "state update" call to complete the defrosting termination control signal issuance. The instruction also serves as an important trigger point for the defrosting control closed loop, associated with the system state update and feedback module of step S600, to achieve complete closed loop management of the defrosting process.

[0047] Step S600 includes at least steps S610-S630: S610, after receiving the termination instruction, obtain the current system operating parameters, perform heat pump mode switching processing based on the communication interface module and the pre-set control logic, and generate a restart configuration; The system first acquires the current system operating parameters through the communication interface module, including but not limited to heat pump operating mode, evaporator temperature, environmental humidity, system pressure, and refrigerant flow rate, and other key indicators. Specifically, the operating parameters are continuously collected by the multi-sensor data acquisition unit and transmitted to the state management module through the real-time data bus, ensuring the time sequence consistency and integrity of the data. After receiving the termination instruction, the system triggers the heat pump mode switching process, executes the switching operation of the heat pump operating mode according to the preset control logic and safety constraints, and combines the current operating parameters. The switching includes switching from the defrosting mode back to the normal heating or cooling mode, and the specific operation involves the cooperative control of the compressor start-stop, four-way valve switching, and fan speed adjustment, etc. During the switching process, the system detects the execution state of each device in real time through the state monitoring module to ensure the accuracy and stability of the switching action. Further, the system dynamically adjusts the switching strategy parameters according to the current environmental conditions and historical operation data, optimizes the response time and energy consumption performance of the switching process. After the heat pump mode switching process is completed, the system generates a system restart configuration, which contains the operating parameter set after switching, initialization instructions, and safety protection parameters, as the basic data for subsequent system state updating and feedback. The system restart configuration is formatted into a standard data structure by the configuration management module, with a timestamp and version number information, to ensure the uniqueness and traceability of the configuration. The system restart configuration, as the output field name of this step, is passed to the "parameter feedback" call of step S620, forming a key data bridge for defrosting termination control and system state updating, and at the same time, the configuration forms a feedback loop with the multi-source data fusion and threshold dynamic adjustment module of step S200, supporting the continuous optimization of dynamic threshold parameters.

[0048] S620, extract key operating parameters from the restart configuration, calculate feedback coefficients based on the configuration analysis module and feedback algorithm model, and generate threshold adjustment suggestions; The system extracts key operating parameters from the configuration, including post-switch heat pump operation mode, evaporator surface temperature, ambient humidity level, system pressure state, and refrigerant flow rate. The key operating parameters are structured by the configuration analysis module to ensure uniform data format and compliance with internal protocol standards. After analysis, the system calculates the feedback coefficient based on the preset feedback algorithm model. The feedback algorithm combines historical operating data, environmental trends, and current system state, using multivariate regression analysis and fuzzy logic reasoning to generate a feedback coefficient for dynamic threshold adjustment. Specifically, the feedback coefficient reflects the influence of the current system operating state on the defrost trigger threshold, quantifying the contribution of environmental and equipment state changes to threshold adjustment. During calculation, the system checks the integrity and validity of the input data, and initiates an exception recording mechanism for abnormal or missing data, saving exception events and data context to ensure the accuracy and reliability of feedback calculation. Further, the feedback coefficient calculation dynamically adjusts feedback model parameters based on seasonal variation characteristics and equipment aging state, achieving personalized customization of threshold adjustment recommendations. The feedback coefficient is output in numerical form, with a timestamp and parameter version information, supporting subsequent time series management and model updates. The threshold adjustment recommendation, as the output field name of this step, is passed to the "closed-loop update" call in step S630, forming a data link between system state update and dynamic threshold adjustment, promoting adaptive optimization of the defrost control strategy.

[0049] S630, write the threshold adjustment suggestion into the dynamic threshold adjustment algorithm parameter library based on the parameter verification module and multi-version management, complete the closed-loop update of the system state; The system writes the adjustment suggestion into the dynamic threshold adjustment algorithm parameter library. Specifically, the parameter library management module performs format checking and version control on the feedback coefficient to ensure that the data meets the storage specifications of the system parameter library. The write operation adopts a transaction mechanism to ensure the atomicity and consistency of parameter library updates and prevent incomplete or conflicting data due to unexpected interruptions. The system further simulates the newly written adjustment suggestion through the parameter verification module to evaluate its impact on the dynamic threshold adjustment algorithm and verify the rationality and safety of the parameter update. After verification, the adjustment suggestion is officially incorporated into the real-time running environment of the dynamic threshold adjustment algorithm, affecting the weight parameter setting and threshold calculation logic in the subsequent step S210 "weighted calculation". The dynamic threshold adjustment algorithm parameter library supports multi-version management and rollback functions to ensure that the system can quickly recover to a stable version in the event of parameter abnormalities. The system regularly checks the integrity and performance of the parameter library and continuously optimizes the parameter update strategy in combination with the device running state and environmental changes. The updated parameter library contains the latest feedback adjustment suggestions and historical version records, with timestamps and version identifiers, supporting real-time calling and adjustment of the dynamic threshold in step S210 "weighted calculation". The updated parameter library serves as the output field name of this step and is passed to the "weighted calculation" in step S210, forming a closed-loop feedback mechanism for the defrosting control system and promoting the continuous self-adaptation and optimization of multi-source data fusion and dynamic threshold adjustment.

Claims

1. A defrosting control method for an air source heat pump, characterized in that, include: Real-time data is acquired from evaporator temperature sensor, ambient humidity sensor, and system pressure sensor. The data is then timestamped based on a synchronous clock, and outlier removal and unit standardization are performed based on threshold range detection, mutation detection, and continuity detection to generate fused input data. Based on the fused input data, temperature, humidity, and pressure feature values ​​are extracted and weighted by a normalization and dynamically adjusted weighting algorithm to generate a frost risk index and a dynamic trigger threshold. A threshold mapping table is then constructed based on a time series format and an incremental update mechanism. The dynamic trigger threshold is obtained from the threshold mapping table, and the frost risk index is compared based on point-by-point comparison and counter mechanism, and the duration is determined based on the sliding time window algorithm to generate a control signal. The system acquires control signals, calculates the rate of change based on numerical differentiation and weighted filtering algorithms, evaluates defrosting effect based on sliding window technology and multi-factor fusion algorithms, generates correction coefficients, adjusts preset defrosting duration parameters based on multiplication or addition models, and generates an optimized termination threshold. Based on the correction coefficient, combined with the optimized termination threshold, the duration is verified using a high-resolution timer and sliding window algorithm, and the state is switched based on the state management module and the preset control strategy to generate a termination command. Upon receiving a termination command, the system obtains system operating parameters, switches the heat pump mode based on the communication interface module and preset control logic, calculates feedback coefficients based on the configuration parsing module and feedback algorithm model, generates restart configuration and threshold adjustment suggestions, and completes closed-loop system status updates based on the parameter verification module and multi-version management.

2. The method according to claim 1, characterized in that, The process of generating fused input data also includes: Real-time data from the evaporator temperature sensor, ambient humidity sensor, and system pressure sensor are acquired, and timestamp alignment is performed based on a synchronous clock to obtain the raw sensor data. Temperature, humidity, and pressure values ​​are extracted from the raw sensor data. Outlier removal and unit standardization are performed based on threshold range detection, mutation detection, and continuity detection to generate cleaned data. The cleaned data is processed using a time window sliding algorithm to synchronize multi-source data and generate fused input data.

3. The method according to claim 1, characterized in that, The process of constructing the threshold mapping table also includes: Temperature, humidity, and pressure features are extracted from the fused input data. A weighted algorithm is then used to calculate the frost risk index based on normalization and dynamic adjustment of weight parameters. Based on the frost risk index, dynamic threshold parameters are calculated using historical frost risk index data and environmental parameters to generate dynamic trigger thresholds. The frost risk index and dynamic trigger threshold are associated and stored based on time series data and an incremental update mechanism to generate a threshold mapping table.

4. The method according to claim 1, characterized in that, The process of generating control signals also includes: Extract the current dynamic trigger threshold from the threshold mapping table, compare the frost risk index based on point-by-point comparison and counter mechanism, and generate a threshold exceedance status marker. Based on the duration of the threshold exceeding the mark, the necessity of defrosting is determined using a sliding time window algorithm and priority rules, and a start command is generated. The startup instructions are prioritized based on a multi-level priority queue management mechanism to generate control signals.

5. The method according to claim 1, characterized in that, The process of generating the optimized termination threshold also includes: The control signal is acquired, and the rate of change is calculated based on the numerical differentiation method and weighted filtering algorithm to generate process indicators. The temperature rise rate and pressure stability value are extracted from the process indicators. The defrosting effect is evaluated based on sliding window technology and multi-factor fusion algorithm, and a correction coefficient is generated. Based on the correction coefficient and using a multiplication or addition model, the preset defrosting duration parameter is adjusted to generate an optimized termination threshold.

6. The method according to claim 1, characterized in that, The process of generating the termination instruction also includes: Obtain the correction coefficient, combine it with the optimized termination threshold, and generate termination candidate markers based on the timestamp alignment algorithm and sliding time window technique; The duration of the termination candidate markers is verified using a high-resolution timer and a sliding window algorithm to generate a qualifying signal. Based on the compliance signal, the system state is switched according to the state management module and the preset control strategy, and a termination command is generated.

7. The method according to claim 1, characterized in that, The process of completing the closed-loop update of system status based on the parameter verification module and multi-version management also includes: After receiving the termination command, the system obtains the current system operating parameters, performs heat pump mode switching based on the communication interface module and preset control logic, and generates a restart configuration. Key runtime parameters are extracted from the restart configuration, and feedback coefficients are calculated based on the configuration parsing module and feedback algorithm model to generate threshold adjustment suggestions. Threshold adjustment suggestions are written into the dynamic threshold adjustment algorithm parameter library based on the parameter verification module and multi-version management to complete the closed-loop update of the system status.

8. The method according to claim 5, characterized in that, The expression for generating the optimized termination threshold also includes: The rate of temperature change was calculated using an improved cross-correlation algorithm. : in, This refers to the rate of temperature change. The index number of the current data point; For summation index variables; The length of the sliding window; This refers to the surface temperature of the evaporator. For timestamps; For delay parameters; This represents the average temperature within the window. This represents the temperature difference between the beginning and end of the window. For time intervals; Calculate the rate of pressure change using the pressure difference: in, This represents the rate of change of pressure. Refrigerant pressure; It is the difference order; Furthermore, a dynamic weighted fusion method is used to evaluate the defrosting effect: in, This is the evaluation value for defrosting effectiveness; These are the weighting coefficients; This represents the historical extreme value of the rate of temperature recovery; This is the stable pressure value; This represents the historical extreme value of pressure stability. Correction coefficients are generated using a sigmoid function: in, This is a correction factor; This is the slope parameter; This serves as the baseline value for the effect; Based on correction factor Adjust the preset defrosting time : in, This is the adjusted defrosting time; Preset defrosting time; To adjust the amplitude coefficient; Build an optimized termination threshold : in, To optimize the termination threshold; The preset threshold; This refers to the sensitivity parameter.

9. The method according to claim 5 or 8, characterized in that, The process of adjusting the preset defrost duration parameters also includes: The preset duration is adjusted by combining the correction coefficient and using a multiplication or addition model. The specific adjustment method is segmented according to the numerical range of the correction coefficient.

10. The method according to claim 5 or 9, characterized in that, The process of generating the optimized termination threshold also includes: Based on the defrosting duration parameter, the optimized termination threshold is calculated. The calculation of the termination threshold takes into account the temperature recovery rate, pressure stability, and changes in ambient humidity, forming a multi-dimensional judgment standard.

Images