Dual-temperature cooperative acquisition circuit, intelligent temperature measurement method and vehicle-mounted refrigerator of vehicle-mounted refrigerator

Abstract

The application belongs to the field of vehicle-mounted refrigerators, and discloses a dual-temperature cooperative acquisition circuit, an intelligent temperature measurement method and a vehicle-mounted refrigerator, which comprise: a synchronous acquisition unit, which is used for synchronously acquiring temperature signals of at least two target regions of the vehicle-mounted refrigerator and outputting corresponding temperature analog signals; an anti-interference processing unit, which is connected with the output end of the synchronous acquisition unit and is used for filtering various interference signals in the vehicle-mounted environment; an adaptive signal conditioning unit, which is connected with the output end of the anti-interference processing unit and is used for adaptively amplifying, filtering and linearizing the temperature analog signals; a main control unit, which is connected with the output end of the adaptive signal conditioning unit and is used for digitizing the temperature signals, performing algorithm operation and logical control; and a wide-adaptive power supply unit, which is connected with the synchronous acquisition unit, the anti-interference processing unit, the adaptive signal conditioning unit and the main control unit and is used for providing a stable working voltage that is adapted to the vehicle-mounted power supply.

Description

Dual-temperature collaborative acquisition circuit, intelligent temperature measurement method and vehicle refrigerator Technical Field

[0001] This manual relates to the field of vehicle refrigerator technology, and in particular to a dual-temperature collaborative acquisition circuit, intelligent temperature measurement method, and vehicle refrigerator. Background Technology

[0002] Temperature acquisition and control for vehicle-mounted refrigerators need to adapt to the complex vehicle environment. Current technology has many key shortcomings, making it difficult to meet the requirements for accurate temperature measurement and stable operation. Specific problems are as follows:

[0003] Firstly, the power supply adaptability and anti-interference capability are insufficient. Most existing vehicle refrigerator power supply modules are simple voltage stabilization designs, which are not compatible with different specifications of vehicle power supplies such as 12V / 24V, and also lack targeted anti-interference structures. They are easily affected by vehicle electromagnetic interference and voltage fluctuations, resulting in unstable power supply, which in turn causes malfunctions in the acquisition circuit and control module. At the same time, when some modules require high-voltage power supply, the matching between the boost module and the acquisition circuit and control module is poor, resulting in large output voltage fluctuations, which affects the working stability of actuators such as defrosting heaters and variable frequency compressors.

[0004] Secondly, the synchronization and anti-interference of multi-zone temperature acquisition are poor. Existing dual-temperature / multi-temperature refrigerators mostly use asynchronous acquisition methods, resulting in time differences in temperature data from different zones. Furthermore, the acquisition link lacks effective signal isolation and anti-interference processing. High-frequency interference and common-mode interference in the vehicle environment can easily lead to temperature signal distortion, making it impossible to achieve accurate and coordinated acquisition of multi-zone temperatures.

[0005] Third, the temperature measurement accuracy and adaptability to operating conditions are insufficient. Traditional temperature measurement methods rely on a single sensor and a single prediction model, lacking an adaptive signal conditioning mechanism, making it difficult to offset sensor nonlinear errors. At the same time, the dynamic compensation dimension is limited, and the impact of key factors such as power supply voltage fluctuations and refrigerator operating status (cooling / heating / standby) is not fully considered. Especially under extreme weather conditions such as high temperature exposure and low temperature freezing, there is a lack of specific correction mechanisms, and the sensor is prone to nonlinear drift, resulting in a significant decrease in temperature measurement accuracy.

[0006] Fourth, the model robustness and data reliability are insufficient. The parameter weights of existing temperature measurement methods are mostly fixed and cannot be dynamically adjusted according to the operating conditions. Furthermore, they lack a multi-dimensional cross-validation mechanism, making it difficult to identify abnormal data. Data errors can easily lead to misjudgments in temperature control strategies, affecting the refrigerator's cooling / heating performance and energy efficiency.

[0007] In summary, existing technologies cannot simultaneously solve the four core problems of "wide-range power supply adaptability, synchronous anti-interference acquisition, accurate temperature measurement under all operating conditions, and high-reliability data verification," and there is an urgent need for a dual-temperature collaborative acquisition circuit and intelligent temperature measurement method for vehicle-mounted refrigerators. Summary of the Invention

[0008] This invention addresses the problems existing in the prior art by proposing a dual-temperature collaborative acquisition circuit, an intelligent temperature measurement method, and an in-vehicle refrigerator. Through multi-unit collaborative design, it achieves synchronous acquisition, anti-interference processing, and adaptive conditioning of temperature signals from multiple regions in an in-vehicle environment. The intelligent temperature measurement method integrates multi-model fusion prediction, multi-dimensional dynamic compensation, extreme weather-specific correction, cross-validation, and parameter weight optimization mechanisms, solving the problems of low temperature acquisition accuracy and poor adaptability under complex in-vehicle operating conditions and extreme weather. It significantly improves the accuracy, stability, and adaptability of temperature acquisition, optimizes the refrigerator's cooling / heating energy efficiency and user experience, and has broad application prospects.

[0009] To achieve the above objectives, this application provides the following technical solution:

[0010] In a first aspect, a dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator includes a synchronous acquisition unit for synchronously acquiring temperature signals from at least two target areas of the vehicle-mounted refrigerator and outputting corresponding temperature analog signals. The synchronous acquisition unit includes at least two temperature sensing components, a synchronous control component, and a signal isolation component. The synchronous control component controls the synchronous start and stop acquisition of each temperature sensing component, and the signal isolation component isolates signal interference between different temperature sensing components. An anti-interference processing unit, with its input connected to the output of the synchronous acquisition unit, is used to filter out various interference signals in the vehicle environment. An adaptive signal conditioning unit, with its input connected to the output of the anti-interference processing unit, is used to adaptively amplify, filter, and linearize the temperature analog signals. A main control unit, with its input connected to the output of the adaptive signal conditioning unit, is used to perform digital processing, algorithm calculation, and logic control on the temperature signals. A wide-adaptive power supply unit, with its output connected to the synchronous acquisition unit, the anti-interference processing unit, the adaptive signal conditioning unit, and the main control unit, is used to provide a stable operating voltage adapted to the vehicle power supply.

[0011] Optionally, the synchronous acquisition unit is connected to the temperature sensing component through the temperature sensor terminal block. The second terminal of the temperature sensor terminal block is connected to resistor one and resistor two respectively. The other end of resistor one is connected to a 5V voltage. Resistor two is connected to the main control unit and capacitor one respectively. The other end of capacitor one is connected to the ground through the first terminal of the temperature sensor terminal block.

[0012] Optionally, the wide-adaptive power supply unit includes an anti-interference module, a sampling and detection module, an intelligent control module, an automatic switching module, and a boost module. One end of the anti-interference module is connected to the vehicle power supply, and the other end is connected to the sampling and detection module and the intelligent control module. The other end of the sampling and detection module is connected to the automatic switching module, which is connected to the core load. The intelligent control module is electrically connected to the sampling and detection module and the automatic switching module. The input end of the boost module is connected to the vehicle power supply, and the output end provides a preset stable voltage for each unit.

[0013] Secondly, the present invention provides an intelligent temperature measurement method for a vehicle-mounted refrigerator, based on the dual-temperature collaborative acquisition circuit of the vehicle-mounted refrigerator described in the first aspect, comprising the following steps:

[0014] S1, System initialization, setting acquisition parameters, model parameters, verification thresholds and extreme weather judgment thresholds, and starting the synchronous acquisition unit;

[0015] S2, Signal preprocessing: The temperature analog signal is converted into a digital temperature value after anti-interference processing and adaptive conditioning;

[0016] S3, Basic data processing, performs extreme value removal and mean calculation on each group of collected temperature digital values ​​to obtain the average digital value;

[0017] S4, based on multi-model fusion temperature prediction, inputs multiple consecutive sets of average numerical values ​​into a preset fusion prediction model and outputs the predicted temperature value;

[0018] S5, multi-dimensional dynamic compensation, corrects the predicted temperature value based on environmental parameters and a preset hierarchical compensation rule to obtain a preliminary corrected temperature value;

[0019] S6, Extreme Weather Specific Correction: The extreme weather judgment model identifies the current weather type and calls the corresponding nonlinear compensation model to perform a second correction on the initial corrected temperature value to obtain the final temperature value.

[0020] S7, multi-dimensional cross-verification, detects the validity of the final temperature value through at least two of the following: regional temperature difference verification, time series trend verification, and sensor consistency verification. If any verification item fails to meet the threshold condition, fault detection is triggered.

[0021] S8, attention weight optimization, adjusts the model parameter weights based on a preset attention mechanism to improve model adaptability to different operating conditions.

[0022] Optionally, in step S3, the basic data processing involves collecting temperature AD values ​​through a temperature sensing component. When the collected temperature AD value is greater than the maximum set AD value, the current AD value is assigned to ADmax; when the collected temperature AD value is less than the minimum set AD value, the current AD value is assigned to ADmin; the collected temperature AD value is recorded, the number of collections is incremented by 1, and the total sum ADsum is calculated; when the number of collections is greater than or equal to 16, the total sum ADsum is subtracted from ADmax and ADmin, the cumulative number of collections is reduced by two accordingly, and the average AD value is calculated by dividing the remaining sum of AD values ​​by the remaining number of collections.

[0023] Optionally, in step S4, the fusion prediction model is an LSTM-RF fusion model, and the specific formula is shown below:

[0024] (1)

[0025] in, To integrate predicted temperature values; Output temperature values ​​for the LSTM model; Output temperature values ​​for the random forest model; For weight fusion.

[0026] Optionally, in step S5, the specific formula corresponding to the hierarchical compensation rule is as follows:

[0027] (2)

[0028] in, To make initial corrections to the temperature value; This refers to the fluctuation in supply voltage. This is the voltage base compensation coefficient; This is a voltage fluctuation correction factor; The operating status coefficients are: cooling = 1, heating = 2, and standby = 0. For state-based compensation coefficients; This is a working status correction factor; Ambient temperature; For reference ambient temperature; This is the environmental compensation coefficient; This is the ambient temperature correction factor.

[0029] Optionally, in step S6, the extreme weather determination model is a multi-feature classification model based on environmental parameters, and the specific formula is shown below:

[0030] (6)

[0031] in, Score based on weather conditions. The weights of the i-th type of environmental parameters; The normalized values ​​of the i-th type of environmental parameters include ambient temperature, humidity, air pressure, and solar radiation intensity. For bias terms; when It was determined to be extreme high-temperature weather; when The weather was determined to be extreme low temperature weather. , This is a preset threshold for judgment.

[0032] Optionally, the nonlinear compensation model for extreme weather is based on an improved Steinhart-Hart equation; the specific formula for compensation in high-temperature extreme weather is shown below:

[0033] (7)

[0034] The specific formula for compensation in extreme low-temperature weather is shown below:

[0035] (8)

[0036] in, , These are the final temperature values ​​under extreme high and low temperatures, respectively. , This is the compensation coefficient for extreme weather conditions. , The attenuation coefficient; , Thresholds for extreme high and low temperatures; A represents the real-time resistance value of the temperature sensor; A, B, and C are the sensor characteristic coefficients.

[0037] Thirdly, the present invention provides a vehicle-mounted refrigerator, comprising a body, the dual-temperature collaborative acquisition circuit described in the first aspect, and an actuator; the body includes at least two storage areas, and the temperature sensing components of the synchronous acquisition unit are respectively disposed in each storage area; the actuator is controlled by the temperature value output by the intelligent temperature measurement method; the actuator includes a variable frequency compressor, a solenoid valve, a temperature control valve, and a defrost heater; the output terminal of the boost module is respectively connected to the defrost drive module and the variable frequency drive module; the main control unit is respectively connected to the defrost drive module, the variable frequency drive module, the display module, and the temperature and humidity detection module; the defrost drive module is connected to the defrost heater; and the variable frequency drive module is connected to the compressor.

[0038] The beneficial effects of this invention are as follows:

[0039] 1. By integrating anti-interference, sampling detection, and voltage boosting and stabilization functions through a wide-adaptability power supply unit, it effectively suppresses vehicle electromagnetic interference and voltage fluctuations, adapts to different vehicle power supplies such as 12V / 24V, ensures stable power of the actuator, and improves power supply reliability and equipment versatility; synchronous acquisition and multi-dimensional anti-interference design, combined with precise circuit connection relationships, improve the synchronization and stability of temperature data, and adapt to complex vehicle interference environments;

[0040] 2. By combining multi-model fusion prediction with multi-dimensional dynamic compensation, along with optimized basic data processing methods, the accuracy of temperature measurement is improved. Simultaneously, in the extreme weather-specific correction mechanism, a nonlinear compensation model based on the improved Steinhart-Hart equation effectively offsets the nonlinear drift of the sensor under extreme temperatures, improving the temperature measurement accuracy to within ±0.3℃. A multi-dimensional cross-validation mechanism enhances the effectiveness and reliability of temperature data. Improved multi-head attention weight optimization strengthens the weight of key parameters under extreme conditions, enhancing the model's adaptability to operating conditions and fault robustness. Attached Figure Description

[0041] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. The same numbers in the drawings denote the same structures or steps.

[0042] Figure 1 is a schematic diagram of the dual-temperature collaborative acquisition circuit of the vehicle refrigerator in Embodiment 1 of this application;

[0043] Figure 2 is a schematic diagram of the anti-interference module circuit of Embodiment 1 of this application as shown in Figure 1;

[0044] Figure 3 is a schematic diagram of the sampling and detection module circuit of Embodiment 1 of this application as shown in Figure 1;

[0045] Figure 4 is a schematic diagram of the boost module circuit of Embodiment 1 of this application as shown in Figure 1;

[0046] Figure 5 is a schematic diagram of the intelligent temperature measurement method for a vehicle-mounted refrigerator according to Embodiment 2 of this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely one preferred embodiment of this application and are only used to explain this application. They do not limit the scope of protection of this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Example 1:

[0049] As shown in Figures 1-4, a dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator includes a synchronous acquisition unit, an anti-interference processing unit, an adaptive signal conditioning unit, a main control unit, and a wide-adaptive power supply unit. The synchronous acquisition unit is used to synchronously acquire temperature signals from at least two target areas of the vehicle-mounted refrigerator and output corresponding temperature analog signals. The anti-interference processing unit, with its input connected to the output of the synchronous acquisition unit, is used to filter out various interference signals in the vehicle environment. The anti-interference processing unit includes at least one filtering component, a transient suppression component, and a common-mode suppression component. The filtering component is used to filter out interference at specific frequencies. The transient suppression component is used to clamp transient abnormal voltages, and the common-mode suppression component is used to suppress common-mode interference. The adaptive signal conditioning unit, with its input connected to the output of the anti-interference processing unit, is used to adaptively amplify, filter, and linearize the temperature analog signal. The main control unit, with its input connected to the output of the adaptive signal conditioning unit, is used to digitally process the temperature signal, perform algorithm calculations, and perform logic control. The wide-adaptive power distribution unit, with its output connected to the synchronous acquisition unit, the anti-interference processing unit, the adaptive signal conditioning unit, and the main control unit, is used to provide a stable operating voltage adapted to the vehicle power supply.

[0050] Specifically, the synchronous acquisition unit includes at least two temperature sensing components, a synchronous control component, and a signal isolation component, which are deployed in each storage area of ​​the vehicle refrigerator. The synchronous control component controls the synchronous start and stop of each temperature sensing component for acquisition, and the signal isolation component is used to isolate signal interference between different temperature sensing components. The acquisition interval of the synchronous acquisition unit is uniformly set to 10-20 milliseconds. The temperature sensing components use thermistor sensors, such as high-precision NTC thermistors. The signal isolation component uses optocouplers or magnetic isolation devices, connected in series between the sensors and the subsequent processing unit to avoid signal crosstalk between different sensors and ensure the independence of the acquired signals. The synchronous acquisition unit is connected to the temperature sensing component through the temperature sensor terminal block. Terminal 2 of the temperature sensor terminal block (e.g., J1) is connected to resistor 1 (R1, 1kΩ) and resistor 2 (R2, 4.7kΩ) respectively. The other end of resistor 1 is connected to a 5V power supply. One end of resistor 2 is connected to the main control unit (e.g., the ADC pin PA0 of STM32F407), and the other end is connected to capacitor 1 (C1, 0.1μF). The other end of capacitor 1 is connected to terminal 1 of the terminal block and then grounded, thus ensuring stable transmission of the sensor signal.

[0051] Specifically, the filtering component of the anti-interference processing unit adopts an LC low-pass filter network. The input of the filtering component is connected to the output of the synchronous acquisition unit to filter out high-frequency interference above 1MHz. The transient suppression component is connected in parallel with a TVS diode to clamp transient high voltage. The common-mode suppression component adopts a differential amplifier circuit. The input is connected to the output of the filtering component. The common-mode rejection ratio is ≥80dB, which effectively suppresses common-mode interference. The output is connected to the input of the adaptive signal conditioning unit.

[0052] Specifically, the adaptive signal conditioning unit includes a programmable amplification module, a Kalman filter module, and a piecewise linearization correction module. The programmable amplification module dynamically adjusts the amplification factor according to the signal strength, with a range of 1-20 times. The Kalman filter module estimates noise in real time and dynamically adjusts the filter coefficient. The piecewise linearization correction module performs linear correction on the temperature range in segments based on the nonlinear characteristics of the sensor, improving the linearity of the signal and ensuring the signal quality input to the main control unit.

[0053] Specifically, the wide-adaptive power distribution unit serves as the energy source for each module, enabling anti-interference, sampling and detection, voltage regulation, and boost processing of the vehicle power supply to ensure stable power supply for the entire system. The wide-adaptive power distribution unit includes an anti-interference module, a sampling and detection module, an intelligent control module, an automatic switching module, and a boost module. One end of the anti-interference module connects to the vehicle power supply (12V / 24V), and the other end connects to the sampling and detection module and the intelligent control module. The other end of the sampling and detection module connects to the automatic switching module, which in turn connects to the core load. The intelligent control module's electrical signals are connected to the sampling and detection module and the automatic switching module. The boost module's input is connected to the vehicle power supply, and its output provides a preset stable voltage to each unit.

[0054] The anti-interference module includes capacitor C1, bidirectional clamping diode D1, diode D2, power switch Q1, resistor R1, power switch Q2, capacitor C2, and capacitor C3. Capacitor C1 and bidirectional clamping diode D1 are connected in parallel with the vehicle input terminal. Diode D2 is connected in parallel with power switch Q2 at the vehicle input terminal. The cathode of diode D2 is connected to the vehicle input terminal, and the anode is connected to the emitter of power switch Q1. The gate of power switch Q1 (NPN transistor, such as S8050) is connected to the power input terminal. The base of the load pump voltage control terminal is connected to one end of resistor R1 (1kΩ), and the other end of resistor R1 is grounded. The collector is connected to the source of power switch Q2 (PMOS transistor, such as AO3401). The drain of Q2 is connected to one end of capacitor C2 (10μF) and capacitor C3 (0.1μF), and the other ends of C2 and C3 are grounded. The cathode of diode D2 (Schottky diode, such as SS34) is connected to the vehicle input terminal, and the anode is connected to the emitter of Q1. This effectively suppresses electromagnetic interference and transient voltage surges.

[0055] The sampling and detection module includes a voltage acquisition module and a current detection module. The voltage acquisition module consists of resistor R3, resistor R2, and capacitor C4. One end of resistor R3 is connected to the anti-interference module, and the other end is connected in series with resistor R2. The other end of resistor R2 is grounded. Capacitor C4 is connected in parallel with resistor R2. The connection point between resistor R3 and resistor R2 is connected to the intelligent control module. The current detection module includes resistor R4 and sampling lines CURRENT_N and CURRENT_P. Resistor R4 is connected in series, and the sampling lines are connected to both ends of R4 and then to the intelligent control module, thereby realizing real-time monitoring of voltage and current.

[0056] The intelligent control module includes an MCU module and a power supply module. The power supply module consists of a voltage regulator chip U1, capacitors C5 and C6. The input terminal of the voltage regulator chip U1 is connected to the anti-interference module, C5 is connected in parallel to the input terminal of U1 and grounded, and the output terminal of U1 is connected in parallel to C6 and grounded and connected to the power interface of the MCU module; thereby realizing intelligent on / off control of the power supply path.

[0057] The automatic switching module includes resistor 7 R5, power switches Q3 and Q4. The two ends of resistor 7 R5 are connected to the source and gate of Q3. The source of Q3 is connected to the anti-interference module, the drain is connected to the core load, the gate is connected to the drain of Q4, the source of Q4 is grounded, and the gate is connected to the control signal output terminal of the MCU module.

[0058] The boost module includes an input detection unit, a control unit, a boost inductor, a boost capacitor, a switching unit, a voltage feedback unit, and an absorption unit. One end of the boost inductor is connected to the input detection unit and the vehicle power supply, and the other end is connected to the boost capacitor through the switching unit. The boost capacitor is connected to each power-consuming unit. The input detection unit is connected to the control unit, and the control unit is connected to the controlled terminal of the switching unit. The voltage feedback unit includes a first voltage divider resistor R75 and a second voltage divider resistor R73 connected in series. One end of the first voltage divider resistor R75 is connected to the boost capacitor, and the other end is connected to one end of the second voltage divider resistor R73 and the feedback terminal of the control unit. The other end of the second voltage divider resistor R73 is grounded. The switching unit includes a first switching transistor 251 and a second switching transistor. 252, one end of which is connected to the other end of the boost inductor; the other end of the first switching transistor 251 is grounded; the other end of the second switching transistor 252 is connected to the boost capacitor; the absorption unit includes an absorption resistor R1 and an absorption capacitor C1 connected in series, with their two ends connected to the two ends of the second switching transistor 252 respectively; the input detection unit includes a first detection resistor R2, a second detection resistor R5, a third detection resistor R6, and a detection capacitor C8. One end of the first detection resistor R2 is connected to the vehicle power supply and one end of the second detection resistor R5, and the other end is connected to the boost inductor and one end of the third detection resistor R6 respectively. The other end of the second detection resistor R5 is connected to one end of the detection capacitor C8 and the control unit respectively. The other end of the third detection resistor R6 is connected to the other end of the detection capacitor C8 and the control unit respectively. The boost capacitor output is divided into multiple paths, which are connected to the defrosting drive module, the main control module, and the frequency converter drive module respectively, providing stable preset voltages such as 36V / 48V to each module, adapting to different vehicle power supplies and ensuring stable power of the actuator.

[0059] Specifically, the main control unit adopts a microcontroller with high-speed computing capabilities, such as the ARM Cortex-M series. The input terminal is connected to the output terminal of the adaptive signal conditioning unit, and at the same time, it communicates bidirectionally with the intelligent control module and boost module control unit of the wide-adaptive power distribution unit to complete signal digital conversion, multi-model algorithm calculation and logic control of each unit, so as to realize the full-link coordination of power supply, acquisition, processing and control.

[0060] Example 2:

[0061] As shown in Figure 5, an intelligent temperature measurement method for a vehicle-mounted refrigerator, based on the dual-temperature collaborative acquisition circuit of the vehicle-mounted refrigerator in Embodiment 1, includes the following steps:

[0062] S1, System initialization, setting acquisition parameters, model parameters, verification thresholds and extreme weather judgment thresholds, and starting the synchronous acquisition unit;

[0063] S2, Signal preprocessing: The temperature analog signal is converted into a digital temperature value after anti-interference processing and adaptive conditioning;

[0064] S3, Basic data processing, performs extreme value removal and mean calculation on each group of collected temperature digital values ​​to obtain the average digital value;

[0065] S4, Multi-model fusion temperature prediction, inputs multiple consecutive sets of average numerical values ​​into a preset fusion prediction model and outputs predicted temperature values;

[0066] S5, multi-dimensional dynamic compensation, corrects the predicted temperature value based on environmental parameters and a preset hierarchical compensation rule to obtain a preliminary corrected temperature value;

[0067] S6, Extreme Weather Specific Correction: The extreme weather judgment model identifies the current weather type and calls the corresponding nonlinear compensation model to perform a second correction on the initial corrected temperature value to obtain the final temperature value.

[0068] S7, multi-dimensional cross-verification, detects the validity of the final temperature value through at least two of the following: regional temperature difference verification, time series trend verification, and sensor consistency verification. If any verification item fails to meet the threshold condition, fault detection is triggered.

[0069] S8, attention weight optimization, adjusts the model parameter weights based on a preset attention mechanism to improve model adaptability to different operating conditions.

[0070] By combining multi-model fusion prediction with multi-dimensional dynamic compensation, along with optimized basic data processing methods, the accuracy of temperature measurement is improved. Simultaneously, a special correction mechanism for extreme weather conditions, based on a nonlinear compensation model using an improved Steinhart-Hart equation, effectively counteracts the nonlinear drift of sensors under extreme temperatures, improving the temperature measurement accuracy to within ±0.3℃. A multi-dimensional cross-validation mechanism enhances the effectiveness and reliability of temperature data. Improved multi-head attention weight optimization strengthens the weight of key parameters under extreme conditions, enhancing the model's adaptability to operating conditions and its fault robustness.

[0071] In a specific embodiment, in step S3, the basic data processing performs extreme value removal and mean calculation on the temperature AD values ​​collected in each group of 16 samples. Specifically, temperature AD values ​​are collected through a temperature sensing component. When the collected temperature AD value is greater than the maximum set AD value, the current AD value is assigned to ADmax; when the collected temperature AD value is less than the minimum set AD value, the current AD value is assigned to ADmin; the collected temperature AD value is recorded, the number of collections is incremented by 1, and the sum ADsum is calculated; when the number of collections is greater than or equal to 16, the sum ADsum is subtracted from ADmax and ADmin, the cumulative number of collections is reduced by 2 accordingly, and the average AD value is calculated by dividing the sum of the remaining AD values ​​by the remaining number of collections; the data collection interval for each group and the output interval of the average AD value are maintained at 180-200 milliseconds to reduce the impact of abnormal data on the measurement results.

[0072] In a specific embodiment, in step S4, the fusion prediction model adopts an LSTM-RF fusion model, combining the temporal correlation capture advantage of the LSTM model with the nonlinear fitting ability of the random forest model. The LSTM model takes four consecutive sets of average numerical values ​​as input, and the hidden layer contains at least two LSTM units. The random forest model takes the same input as the LSTM model, and the number of decision trees is 50-100. Through the fusion prediction model, the predicted temperature value is output, improving the prediction accuracy. The specific formula is shown below:

[0073] (1)

[0074] in, To integrate predicted temperature values; Output temperature values ​​for the LSTM model; Output temperature values ​​for the random forest model; To merge weights, and ;in, Predict the mean squared error for the LSTM model; The mean squared error is predicted for the random forest model, and the fusion weights are dynamically allocated based on the model accuracy.

[0075] In a specific embodiment, in step S5, based on multi-dimensional parameters such as power supply voltage fluctuations, operating status, and ambient temperature, the predicted temperature value is corrected using a hierarchical compensation formula to obtain a preliminary corrected temperature value. Correction factors for each dimension are introduced, and the compensation weights are dynamically adjusted according to actual operating conditions to ensure the accuracy of temperature measurement under normal operating conditions. The specific formula corresponding to the hierarchical compensation rule is shown below:

[0076] (2)

[0077] in, To make initial corrections to the temperature value; This refers to the fluctuation in supply voltage. This is the voltage base compensation coefficient; This is a voltage fluctuation correction factor; The operating status coefficients are: cooling = 1, heating = 2, and standby = 0. For state-based compensation coefficients; This is a working status correction factor; Ambient temperature; For reference ambient temperature; This is the environmental compensation coefficient; This is the ambient temperature correction factor.

[0078] Voltage fluctuation correction factor Working status correction factor and ambient temperature correction factor They are respectively:

[0079] (3)

[0080] (4)

[0081] (5)

[0082] in, Voltage influence coefficient; This represents the environmental impact coefficient.

[0083] In a specific embodiment, in step S6, the type of extreme weather is determined using a multi-feature classification model. This model integrates parameters such as ambient temperature, humidity, air pressure, and solar radiation intensity, and obtains a determination score through normalization and weighted summation. When the score exceeds the high-temperature threshold, the high-temperature extreme weather compensation model is invoked; when the score is below the low-temperature threshold, the low-temperature extreme weather compensation model is invoked. The compensation model is based on an improved Steinhart-Hart equation, fully considering the nonlinear drift characteristics of sensors under extreme temperatures, and dynamically adjusts the compensation intensity through an exponential decay function to achieve high-precision secondary correction. Specifically, the multi-feature classification model based on environmental parameters has the following formula:

[0084] (6)

[0085] in, Score based on weather conditions. The weights of the i-th type of environmental parameters; The normalized values ​​of the i-th type of environmental parameters include ambient temperature, humidity, air pressure, and solar radiation intensity. For bias terms; when It was determined to be extreme high-temperature weather; when The weather was determined to be extreme low temperature weather. , This is a preset threshold for judgment.

[0086] The specific formula for compensation in extreme high-temperature weather is shown below:

[0087] (7)

[0088] The specific formula for compensation in extreme low-temperature weather is shown below:

[0089] (8)

[0090] in, , These are the final temperature values ​​under extreme high and low temperatures, respectively. , This is the compensation coefficient for extreme weather conditions. , The attenuation coefficient; , Thresholds for extreme high and low temperatures; A represents the real-time resistance value of the temperature sensor; A, B, and C are the sensor characteristic coefficients.

[0091] In one specific embodiment, in step S7, multi-dimensional cross-validation comprehensively detects the validity of temperature data through at least two of the following: regional temperature difference verification, time-series trend verification, and sensor consistency verification. If any verification item exceeds the threshold, a fault detection process is triggered, including re-acquisition, signal path detection, sensor status detection, and alarm signal output. Specifically, the threshold for regional temperature difference verification is 5-10℃; the specific formula is shown below:

[0092] (9)

[0093] in, This represents the temperature difference between different storage areas of the vehicle refrigerator. This is the corrected final temperature value for the first storage area; This is the final temperature value for the second storage area after correction.

[0094] Time-series trend verification, with a threshold of 2℃; the specific formula is shown below:

[0095] (10)

[0096] in, This represents the trend deviation of temperatures collected in two consecutive measurements within the same area. This is the final temperature value after correction for the current data acquisition cycle; This is the final temperature value after correction from the previous data collection cycle; This is the preset allowable temperature fluctuation value between adjacent acquisition cycles.

[0097] Sensor consistency verification, threshold is 3 The specific formula is shown below:

[0098] (11)

[0099] in, The standard deviation of the AD values ​​collected continuously m times by the same temperature sensor represents the dispersion of the collected data. This represents the number of consecutive data collections, and its value is a positive integer. This represents the original AD value of the temperature obtained from the i-th acquisition. This represents the average value of the AD values ​​collected consecutively over m periods. The standard deviation of the AD values ​​collected under the historical normal operating conditions of the sensor is used as the benchmark dispersion index.

[0100] In a specific embodiment, step S8 employs an improved multi-head attention mechanism, setting up 2-4 independent attention heads, each trained based on different parameter association dimensions. In the actual workflow of the vehicle-mounted refrigerator, the model parameter weights are dynamically adjusted by calculating the multi-head attention score, explicitly strengthening the weight proportion of key parameters under extreme weather conditions, and improving the model's adaptability and fault robustness to extreme operating conditions. The specific formula for weight adjustment is shown below:

[0101] (12)

[0102] in, The corrected weights for the i-th type of parameters; Basic weights; The multi-head attention score for the i-th type of parameters; The sum of the multi-head attention scores for all parameters.

[0103] Specifically, the formula for calculating the multi-head attention score is as follows:

[0104] (13)

[0105] Where k is the sequence number of the 3 welfare attention heads; Dimensions for each attention head; This is the output weight matrix for the k-th head; , , These are the query, key, and value matrices for the k-th head, respectively.

[0106] Example 3:

[0107] A vehicle-mounted refrigerator, integrating the intelligent temperature measurement method of the vehicle-mounted refrigerator of Embodiment 2, includes a body, a dual-temperature collaborative acquisition circuit as in Embodiment 1, a display module, a control module, and an actuator. The actuator operates based on the temperature value output by the intelligent temperature measurement method. The actuator includes a variable frequency compressor, a solenoid valve, a temperature control valve, and a defrost heater. The output of a boost module is connected to a defrost drive module and a variable frequency drive module, respectively, providing a stable power supply to the actuator. The defrost drive module includes a relay; the output of the boost module is connected to the relay contacts; the main control unit is connected to the relay coil, the variable frequency drive module, the display module, and the temperature and humidity detection module; the defrost drive module is connected to the defrost heater, controlling its operation by controlling the on / off state of the relay contacts; the variable frequency drive module is connected to the compressor, and the main control unit controls the compressor speed by adjusting the output frequency of the variable frequency drive module.

[0108] The main control unit is bidirectionally connected to the display module (display control circuit board + display screen), the control module (user command input), and the temperature and humidity detection module to realize temperature information display, user command reception, and environmental parameter acquisition. The main control unit dynamically adjusts the working status of the actuator based on the accurate temperature value output by the intelligent temperature measurement method and the environmental parameters, so as to realize independent temperature control and intelligent defrosting in multiple areas and optimize cooling / heating energy efficiency.

[0109] In extreme weather conditions, the control strategy is dynamically adjusted based on the specially modified temperature value, and the wide-adaptability power supply unit ensures stable power output, thus jointly improving the refrigerator's ability to adapt to extreme environments.

[0110] The above-described specific embodiments are preferred embodiments of the dual-temperature collaborative acquisition circuit, intelligent temperature measurement method, and vehicle refrigerator of this application. They are not intended to limit the specific scope of this application. The scope of this application includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape and structure of this application are within the protection scope of this application.

Claims

1. A dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator, characterized in that, include: A synchronous acquisition unit is used to synchronously acquire temperature signals from at least two target areas of the vehicle-mounted refrigerator and output corresponding temperature analog signals. The synchronous acquisition unit includes at least two temperature sensing components, a synchronous control component, and a signal isolation component. The synchronous control component controls each temperature sensing component to start and stop acquiring synchronously, and the signal isolation component is used to isolate signal interference between different temperature sensing components. An anti-interference processing unit, with its input end connected to the output end of the synchronous acquisition unit, is used to filter out various interference signals in the vehicle environment; An adaptive signal conditioning unit, with its input connected to the output of the anti-interference processing unit, is used to adaptively amplify, filter, and linearize the temperature analog signal. A main control unit, with its input connected to the output of the adaptive signal conditioning unit, is used to perform digital processing, algorithm calculations, and logic control on the temperature signal. A wide-adaptive power supply unit, with its output connected to the synchronous acquisition unit, the anti-interference processing unit, the adaptive signal conditioning unit, and the main control unit, is used to provide a stable operating voltage compatible with the vehicle power supply. The system also includes an intelligent temperature measurement method for implementing a dual-temperature collaborative acquisition circuit, comprising the following steps: S1, system initialization, setting acquisition parameters, model parameters, verification thresholds, and extreme weather judgment thresholds, and starting the synchronous acquisition unit; S2, signal preprocessing, converting the temperature analog signal into digital temperature values ​​after anti-interference processing and adaptive conditioning; S3, basic data processing, performing extreme value removal and... S4. Calculate the mean value to obtain the average numerical value; S5. Based on multi-model fusion temperature prediction, input multiple consecutive sets of average numerical values ​​into a preset fusion prediction model and output the predicted temperature value; S6. Perform multi-dimensional dynamic compensation, correcting the predicted temperature value based on environmental parameters using preset hierarchical compensation rules to obtain a preliminary corrected temperature value; S7. Perform extreme weather-specific correction, identifying the current weather type through an extreme weather determination model and calling the corresponding nonlinear compensation model to perform a secondary correction on the preliminary corrected temperature value to obtain the final temperature value; S8. Perform multi-dimensional cross-validation, detecting the validity of the final temperature value through at least two of the following: regional temperature difference verification, time-series trend verification, and sensor consistency verification. If any verification item fails to meet the threshold condition, fault detection is triggered; S9. Optimize attention weights, correcting the model parameter weights based on a preset attention mechanism to improve the model's adaptability to operating conditions. The specific formulas corresponding to the hierarchical compensation rules in step S5 are shown below: (2) Among them, To make initial corrections to the temperature value; This refers to the fluctuation in supply voltage. This is the voltage base compensation coefficient; This is a voltage fluctuation correction factor; The operating status coefficients are: cooling = 1, heating = 2, and standby = 0. For state-based compensation coefficients; This is a working status correction factor; Ambient temperature; For reference ambient temperature; This is the environmental compensation coefficient; This is an ambient temperature correction factor. To integrate predicted temperature values.

2. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, The synchronous acquisition unit is connected to the temperature sensing component through the temperature sensor terminal block. Terminal 2 of the temperature sensor terminal block is connected to resistor 1 and resistor 2 respectively. The other end of resistor 1 is connected to a 5V voltage. Resistor 2 is connected to the main control unit and capacitor 1 respectively. The other end of capacitor 1 is connected to ground through terminal 1 of the temperature sensor terminal block.

3. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, The wide-adaptive power supply unit includes an anti-interference module, a sampling and detection module, an intelligent control module, an automatic switching module, and a boost module. One end of the anti-interference module is connected to the vehicle power supply, and the other end is connected to the sampling and detection module and the intelligent control module. The other end of the sampling and detection module is connected to the automatic switching module, which is connected to the core load. The intelligent control module is electrically connected to the sampling and detection module and the automatic switching module. The input end of the boost module is connected to the vehicle power supply, and the output end provides a preset stable voltage to each unit.

4. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, In step S3, the basic data processing involves collecting temperature AD values ​​through a temperature sensing component. When the collected temperature AD value is greater than the maximum set AD value, the current AD value is assigned to ADmax; when the collected temperature AD value is less than the minimum set AD value, the current AD value is assigned to ADmin. The collected temperature AD value is recorded, the number of collections is incremented by 1, and the total sum ADsum is calculated. When the number of collections is greater than or equal to 16, the total sum ADsum is subtracted from ADmax and ADmin, the cumulative number of collections is reduced by two accordingly, and the average AD value is calculated by dividing the remaining sum of AD values ​​by the remaining number of collections.

5. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, In step S4, the fusion prediction model is an LSTM-RF fusion model, and the specific formula is as follows: (1) Among them, Output temperature values ​​for the LSTM model; Output temperature values ​​for the random forest model; For weight fusion.

6. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, In step S6, the extreme weather determination model is a multi-feature classification model based on environmental parameters, and the specific formula is as follows: (6) Among them, Score based on weather conditions. The weights of the i-th type of environmental parameters; The normalized values ​​of the i-th type of environmental parameters include ambient temperature, humidity, air pressure, and solar radiation intensity. For bias terms; when It was determined to be extreme high-temperature weather; when The weather was determined to be extreme low temperature weather. 、 This is a preset threshold for judgment.

7. The dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator according to claim 1, characterized in that, The nonlinear compensation model for extreme weather is based on an improved Steinhart-Hart equation; the specific formula for compensation in high-temperature extreme weather is shown below: (7) Compensation for extreme low-temperature weather, the specific formula is as follows: (8) Among them, 、 These are the final temperature values ​​under extreme high and low temperatures, respectively. 、 This is the compensation coefficient for extreme weather conditions. 、 The attenuation coefficient; 、 Thresholds for extreme high and low temperatures; A represents the real-time resistance value of the temperature sensor; A, B, and C are the sensor characteristic coefficients.

8. A vehicle-mounted refrigerator, characterized in that, The device includes a body, a dual-temperature collaborative acquisition circuit for a vehicle-mounted refrigerator as described in claim 3, and an actuator; the body includes at least two storage areas, and the temperature sensing components of the synchronous acquisition unit are respectively disposed in each storage area; the actuator is controlled by the temperature value output by the intelligent temperature measurement method; the actuator includes a variable frequency compressor, a solenoid valve, a temperature control valve, and a defrost heater; the output terminal of the boost module is connected to the defrost drive module and the variable frequency drive module respectively; the main control unit is connected to the defrost drive module, the variable frequency drive module, the display module, and the temperature and humidity detection module respectively; the defrost drive module is connected to the defrost heater; and the variable frequency drive module is connected to the compressor.

Images