A smart temperature-controlled cooling infrared detector

Abstract

This invention belongs to the field of infrared detection technology, and particularly relates to an intelligent temperature-controlled cooling infrared detector. The invention comprises a type II superlattice focal plane array chip, a Stirling cooler, a temperature sensing module, an intelligent control module, and a mode switching interface, forming a closed-loop temperature control system. It collects multi-dimensional operating parameters and calculates the scenario demand index and joint optimization objective. Kalman filtering is used to achieve parameter denoising and trend prediction. It adaptively switches between long-life, high-sensitivity operating modes and completes closed-loop temperature control of the chip through smooth PWM adjustment and temperature transition trajectory planning. Compared with existing technologies, this invention solves the pain point of the inability to simultaneously achieve high detection sensitivity and cooler lifespan in infrared detectors, achieving hysteresis-free mode switching and temperature overshoot-free control, effectively balancing detection performance and equipment lifespan, and improving the detector's stability and service life under all operating conditions.

Description

Technical Field

[0001] This invention relates to the field of infrared detection technology, and more specifically, to an intelligent temperature-controlled cooling infrared detector. Background Technology

[0002] Cooled infrared detectors are key functional components in infrared imaging and target detection systems, and their performance directly determines the system's target detection accuracy, detection range, and long-term operational reliability. In particular, cooled infrared detectors based on type-II superlattice (T2SL) materials possess significant technological advantages over traditional mercury cadmium telluride (MCT) detectors, including higher material stability, stronger compatibility with semiconductor processes, good mass production capabilities, and a wide range of cooling temperature adaptability. Therefore, type-II superlattice cooled infrared detectors have become the mainstream development direction for next-generation high-performance cooled infrared detectors.

[0003] However, existing Type II superlattice cooled infrared detectors generally employ a fixed temperature control scheme or only support manual mode switching, failing to intelligently adjust the chip temperature according to the dynamic requirements of actual application scenarios. In engineering applications, this fixed control method suffers from a technical deficiency that cannot balance detection sensitivity and the lifespan of the cooler. Specifically, this manifests in the following aspects:

[0004] Performance degradation in long-cycle missions: In long-cycle missions such as extended cruise surveillance, a fixed low operating temperature causes the Stirling cooler to operate under high load for extended periods, significantly shortening its lifespan. For example, if the chip temperature drops from 120K to 100K, the cooler's lifespan will decrease from 8000 hours to 6000 hours. Conversely, a fixed high operating temperature leads to increased dark current in the detector chip and deterioration of the noise equivalent temperature difference (NETD), failing to meet the stringent performance requirement of ≤15mK and impacting target identification and detection accuracy.

[0005] Lack of scene adaptability: Existing temperature control solutions lack scene adaptability. Most solutions rely on manual commands for mode switching, which is slow to respond and complex to operate. They cannot respond in a timely manner in dynamic scenarios such as patrol monitoring and sudden weak target identification, resulting in detection performance that cannot match mission requirements in real time.

[0006] Limitations of control algorithms: Current temperature control algorithms are mostly based on simple temperature feedback mechanisms and employ fixed-temperature control methods. They fail to effectively integrate multi-dimensional information such as target signal strength, external command frequency, and historical mission data, and lack predictive control capabilities. Consequently, the system cannot anticipate changes in scenario requirements, nor can it construct a joint optimization target for refrigerator lifespan and detection sensitivity. This results in the detector often operating in a state of optimal performance or lifespan in complex mission scenarios, failing to achieve optimal overall performance throughout the mission cycle.

[0007] Therefore, there is an urgent need for a new type of cooled infrared detector that can intelligently adjust the chip's operating temperature according to the dynamic requirements of actual application scenarios. This detector needs to achieve automatic and smooth switching between long-life mode and high-sensitivity mode through a multi-parameter fusion adaptive predictive control algorithm, adapting to complex and ever-changing infrared detection application scenarios, balancing detection performance and cooler lifespan, and improving the overall efficiency and reliability of the system. Summary of the Invention

[0008] In view of the shortcomings of the existing technology, the purpose of this application is to provide an intelligent temperature-controlled cooling infrared detector.

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

[0010] A smart temperature-controlled cooling infrared detector, characterized in that it comprises a superlattice chip, a cooler, a temperature sensing module, an intelligent control module, and a mode switching interface, wherein:

[0011] The superlattice chip is mounted on the cold head of the refrigerator and is used to detect the infrared radiation of the target and output the target signal.

[0012] The refrigeration unit has its PWM input terminal directly connected to the intelligent control module and is used to cool the superlattice chip;

[0013] The temperature sensing module outputs a signal that is connected to the intelligent control module to collect temperature information from the superlattice chip and the cold head of the refrigerator.

[0014] The mode switching interface is connected to the intelligent control module via a communication bus, and is used to receive external task commands and transmit them to the intelligent control module.

[0015] The intelligent control module, built on a microprocessor, receives various parameter information from the temperature sensing module and the mode switching interface, executes a multi-parameter fusion adaptive predictive control algorithm, and outputs a PWM control signal to control the operating temperature of the refrigerator. The intelligent control module has at least two operating modes: a long-life mode and a high-sensitivity mode.

[0016] In a preferred embodiment, the superlattice chip is a type II superlattice focal plane array chip; the refrigerator is an integrated Stirling refrigerator.

[0017] In a preferred embodiment, the intelligent control module includes a data acquisition unit for acquiring chip temperature, temperature change rate, target signal strength, external command frequency, and historical task information of the temperature sensing module and mode switching interface.

[0018] In a preferred embodiment, the execution of the multi-parameter fusion adaptive predictive control algorithm includes the following operations:

[0019] The scenario demand index is calculated by weighting and fusing various parameters based on the temperature change rate, target signal strength, external command frequency, and historical mission change trends to obtain a scenario demand index that represents the comprehensive requirements of the current mission for detection sensitivity and refrigerator lifespan.

[0020] The joint optimization target calculation combines the predicted lifespan of the chiller with the estimated detection sensitivity, assigns adjustable lifespan weights and sensitivity weights respectively, and performs weighted calculations to obtain the joint optimization target value. The current target working mode is then determined based on the optimization target value.

[0021] Kalman filtering prediction performs filtering and trend prediction on the scenario demand index and joint optimization target value within a preset time window, and determines the temperature transition trajectory and mode switching time based on the prediction results.

[0022] In a preferred embodiment, the adjustment of lifetime weight and sensitivity weight in the joint optimization objective calculation satisfies the following rules: when the remaining task duration exceeds a preset duration threshold, the lifetime weight is increased to maintain the refrigerator operating at a higher operating temperature; when the external command frequency exceeds a preset frequency threshold or the target signal strength is lower than a preset strength threshold, the sensitivity weight is increased to enable the refrigerator to operate at a lower operating temperature.

[0023] In a preferred embodiment, the intelligent control module performs control by including the following steps:

[0024] Power on and initialize, then enter the default long life mode, and set the corresponding target operating temperature range and control parameters of the PWM control signal.

[0025] Periodically collect chip temperature, temperature change rate, target signal strength, external command frequency, and historical task trend parameters;

[0026] Determine the current target working mode and target temperature adjustment strategy based on the scenario demand index and joint optimization target value;

[0027] Based on the prediction results of the Kalman filter, the control parameters of the PWM control signal are adjusted so that the operating temperature of the refrigerator gradually approaches the target temperature along the preset transition trajectory.

[0028] Repeat the above steps of data collection, calculation, prediction, and adjustment until the system is powered off or a stop command is received.

[0029] In a preferred embodiment, the intelligent control module receives task instructions sent by an external device through a mode switching interface. The task instructions include task type, task priority, and task duration.

[0030] The intelligent control module adjusts the weight parameters of the scenario demand index and the lifetime and sensitivity weights of the optimization target according to the task instructions, so as to realize the automatic switching between long lifetime mode and high sensitivity mode.

[0031] In a preferred embodiment, the intelligent control module predicts that the target signal strength will drop to a preset threshold within a preset time window based on the Kalman filter, and switches from the long life mode to the high sensitivity mode in advance, so that the refrigerator completes the temperature transition before the target signal strength actually drops.

[0032] In a preferred embodiment, the intelligent control module controls the rate of change of the refrigerator's operating temperature based on the temperature transition trajectory predicted by the Kalman filter, so that the operating temperature gradually approaches the target temperature along the temperature transition trajectory, thereby suppressing temperature fluctuations during mode switching.

[0033] A computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it is used to implement the intelligent temperature control function of an intelligent temperature-controlled cooling infrared detector.

[0034] By adopting the above technical solution, the beneficial effects of the present invention are as follows:

[0035] This invention solves the technical contradiction between the detection sensitivity of traditional infrared detectors and the lifespan of the refrigerator by using an intelligent dual-mode switching technology consisting of a multi-parameter fusion adaptive predictive control algorithm, Kalman filtering trend prediction, and PWM smooth adjustment. It achieves a dynamic balance between detection performance and equipment lifespan. The system has efficient scene adaptation capability. By collecting multi-dimensional parameters such as temperature, target signal, external commands, and historical tasks, and calculating the scene demand index, it can automatically identify task requirements and complete mode switching to achieve stable and accurate closed-loop temperature control. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the system hardware structure in one embodiment of this application;

[0037] Figure 2This is a schematic diagram of the functional units of the intelligent control module in one embodiment of this application;

[0038] Figure 3 This is a schematic diagram of the intelligent temperature control method in one embodiment of this application. Detailed Implementation

[0039] In the following description, many technical details are presented to help the reader better understand this application. However, those skilled in the art will understand that the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments.

[0040] To make the purpose, technical solution, and advantages of this application clearer, the following will be described in conjunction with the appendix. Figure 1-3 The embodiments of this application will be described in further detail.

[0041] like Figure 1 As shown, the intelligent temperature-controlled cooling infrared detector in this embodiment is modularly integrated from five core components. These components form a closed-loop temperature control system through physical bonding and standardized electrical connections. The specific implementation and connection relationships are as follows:

[0042] For infrared detection, a Type II superlattice focal plane array (SLA) chip is preferred, specifically a 640×512 mid-wave infrared T2SL SLA chip with a pixel pitch of 15μm and a spectral response range of 3~5μm. The chip is vacuum-bonded to the cold finger tip of the integrated Stirling refrigerator using a high thermal conductivity oxygen-free copper heat sink (thermal conductivity ≥400W / (m・K)) to achieve efficient thermal coupling. The chip readout circuit is connected to the package lead pad via gold wire bonding, and its target signal output channel is multiplexed to the GPIO digital sampling interface of the intelligent control module, outputting normalized target signal intensity data.

[0043] The integrated Stirling chiller uses a commercial rotary integrated Stirling chiller, supporting a 24VDC power supply. It requires a rated cooling capacity of ≥1.5W at 100K and a minimum cooling temperature of ≤80K. The chiller's PWM control terminal is directly connected to the PWM output interface of the intelligent control module via a high-frequency shielded cable, receiving an adjustable duty cycle PWM signal to linearly adjust the cooling capacity. The chiller's cold head and chip heat sink are integrated and bonded together, and the cold fingers are vacuum-sealed with stainless steel to reduce heat leakage.

[0044] The temperature sensing module employs two high-precision PT1000 platinum resistance temperature sensors, with an optimal measurement accuracy of ±0.3K. The sensors are mounted on the non-photosensitive area of ​​the chip and the contact surface of the refrigerator's cold head, simultaneously acquiring the temperatures of both the chip and the cold head. The analog temperature signal output from the temperature sensing module is processed by a signal conditioning circuit and then input to the 16-bit ADC acquisition interface of the intelligent control module, outputting temperature data and the rate of temperature change.

[0045] The intelligent control module, as the core control hub of the system, is built on the high-performance STM32H743 series microprocessor. The electrical connections between the intelligent control module and other components are as follows: the ADC interface connects to the temperature sensing module, the GPIO digital sampling interface connects to the target signal output terminal of the chip, the CAN bus interface connects to the mode switching interface, the PWM output interface connects to the PWM input terminal of the refrigerator, the CAN bus interface connects to the mode switching interface, and the power interface shares a 24VDC input with the refrigerator (which is stepped down to 3.3V / 5V by the DC-DC module to power the module).

[0046] The mode switching interface uses a 9-pin D-type connector and connects to the CAN communication interface of the intelligent control module via the CAN bus to achieve bidirectional communication, including: receiving task instruction frames sent by the external system (e.g., instruction ID: 0x123, the data segment contains task type, priority, remaining duration, and mode switching command) and feeding back detector status frames to the external system (e.g., status ID: 0x456, the data segment contains current operating mode, chip temperature, refrigerator PWM duty cycle, estimated NETD value, and predicted lifetime value).

[0047] like Figure 2 As shown, the intelligent control module incorporates a multi-parameter fusion adaptive predictive control algorithm. This algorithm is implemented using C language programming and is embedded in the microprocessor's on-chip Flash memory. It consists of five functional units, each interacting with the other in real-time via shared on-chip memory. The specific implementation is as follows:

[0048] Data acquisition unit, with a fixed acquisition cycle The specific steps for performing multi-dimensional parameter acquisition and preprocessing are as follows:

[0049] Temperature parameter acquisition and rate of change calculation: Acquire the current chip temperature output by the temperature sensing module. Cold head temperature ;Calculate the temperature change rate based on the chip temperature difference between adjacent acquisition cycles. (In the formula, This is the chip temperature at the previous data acquisition moment. (For fixed data collection period).

[0050] Core detection parameter acquisition, acquisition chip output target signal strength (After hardware circuit normalization, the value range is) Frequency of external commands sent by the acquisition mode switching interface (After linear transformation and normalization, the range of values ​​is) ).

[0051] Historical data is collected and fitted, reading historical mode switching records and cumulative chiller running time from external storage units. The trend of task requirement changes is obtained through linear fitting. The range of values ​​after fitting is (The larger the value, the stronger the mission's need for high-sensitivity detection.)

[0052] Data preprocessing includes: moving average filtering, and applying a window length to all collected parameters. Moving average filtering eliminates noise; the filtering formula is as follows: (In the formula: The parameters to be filtered include , , , , (The filtered parameter values); outlier removal, setting the temperature parameter outlier threshold to... ,like If the chip temperature is the average of the last 10 acquisition cycles, then the outlier is removed and the valid temperature value from the previous cycle is used; normalization is performed, and all preprocessed parameters are mapped to the [0,1] interval using min-max normalization, with the normalization formula being... (In the formula: , (These are the physical extreme values ​​of each parameter). After preprocessing, all normalized parameters are output to the scenario demand index calculation unit.

[0053] The scenario demand index calculation unit calculates the scenario demand index S in real time through weighted fusion operations based on multi-dimensional normalized parameters after data preprocessing, and simultaneously outputs S to the joint optimization objective calculation unit and the Kalman filter prediction unit. Specifically, the core formula for calculating the scenario demand index S is:

[0054]

[0055] The parameter definition and adjustment rules in the formula are as follows: initial weight configuration (Target signal weight) (Weight of temperature change rate) (External instruction frequency weight) and (Historical trend weight), and satisfying ; 0.01 is to avoid The calculation error that occurs when the value is 0 is a standard error prevention measure in engineering implementation; the dynamic weight adjustment rule is based on the real-time power consumption of the chiller. Time adjustment to , (With other weights remaining unchanged), when the external instruction marks the task priority as "high sensitivity", adjust to , (The remaining weights remain unchanged); The range of values ​​is The larger the value, the stronger the current mission's requirement for high-sensitivity detection by the detector.

[0056] The joint optimization objective calculation unit constructs an optimization objective function with refrigerator lifespan and detection sensitivity as its two core components, calculates the joint optimization objective value J, and outputs J to the Kalman filter prediction unit and the PWM output execution unit. Simultaneously, it receives prediction correction parameters from the Kalman filter prediction unit and iteratively optimizes the weight configuration. Specifically, the core formula for calculating the joint optimization objective value J is:

[0057]

[0058] The parameter definitions and weight adjustment rules in the formula are as follows: The normalized predicted remaining lifespan of the chiller (calculated using a linear extrapolation model based on the current PWM duty cycle and cumulative running time, with a range of values). (A higher value indicates a longer remaining lifespan for the refrigeration unit). The normalized detector noise equivalent temperature difference (NETD) estimate is calculated based on the chip temperature and dark current model (i.e.: , , These are the model fitting coefficients. Range of values ​​after normalization (A smaller value indicates higher detector sensitivity); weighting configuration For lifespan weight, For sensitivity weights, and satisfying When external instructions indicate the remaining time of the task Time adjustment to , ,when times / minute or Time adjustment to , .

[0059] The Kalman filter prediction unit considers the scenario demand index S, the joint optimization objective value J, and the chip temperature. Target signal strength Filtering and denoising, along with trend prediction, are the core components for achieving "hysteresis-free mode switching and overshoot-free temperature control." Specifically, based on the parameter acquisition characteristics of the infrared detector, system operating rules, and actual interference types, core state equations and observation equations adapted to the temperature control scenario of this infrared detector are constructed:

[0060]

[0061]

[0062] Where: state vector It contains all the key parameters that determine mode switching and temperature control; the state transition matrix. Observation matrix All use identity matrices to adapt to the system’s short 10ms acquisition cycle and reflect the actual physical law that parameters remain approximately unchanged over a short period of time. , The covariance matrices of the system noise and the measurement noise are respectively. , The parameters are set differently based on the degree of interference they are subjected to, with the chip temperature having a higher noise weight than other parameters, which is consistent with the actual acquisition characteristics. The measured parameters output by the data acquisition unit are used as the observation input for filtering.

[0063] Furthermore, the state equation provides the "theoretical prediction value," while the observation equation provides the "noisy measured value." Kalman filtering, by solving the two equations, performs a weighted fusion of the "predicted value + measured value," ultimately obtaining the denoised, accurate true value and the trend prediction value for future times. , , This directly serves as the core basis for the PWM output execution unit mode decision, achieving "predicting scenario requirements in advance and switching modes without lag." This embodiment uses a discrete Kalman filter algorithm for simplified engineering solutions, retaining only the core "prediction step—update step" process, which can be directly compiled into C language code for implementation.

[0064] The PWM output execution unit receives the trend prediction value output by the Kalman filter prediction unit. , Temperature transition trajectory, execution mode decision and closed-loop temperature control operation:

[0065] Perform the following operations:

[0066] like or , ( This refers to jointly optimizing the target threshold. Switch to high sensitivity mode and set the target temperature. PWM duty cycle like and Maintain / switch back to long life mode, set target temperature. PWM ;like If the target signal strength is preset to a threshold, then the high sensitivity mode switching will be triggered directly;

[0067] Furthermore, the current PWM duty cycle adjustment step size is calculated based on the temperature transition trajectory (single adjustment step size ≤ 5%), and a corresponding PWM control signal for the duty cycle is generated and output to the PWM input terminal of the chiller to achieve closed-loop precise control of the chip temperature (steady-state temperature control accuracy ±0.3K).

[0068] like Figure 3 As shown, the intelligent temperature control method in this embodiment is executed cyclically by the intelligent control module, forming a complete closed loop of "initialization → data acquisition → calculation → prediction → decision-making → temperature control → cycle". The specific steps are as follows:

[0069] Step S1, after the system is powered on:

[0070] The microprocessor completes peripheral initialization (ADC, PWM, CAN, GPIO), memory unit initialization (EEPROM), and loads default control parameters: initial weights of the scenario demand index. , , , ), pattern decision threshold ( ,

[0071] );

[0072] By default, it switches to long life mode and sets the target temperature. PWM duty cycle ;

[0073] The driver starts the refrigerator to cool the device and controls the chip temperature. It stabilized at 130K±0.3K within 60 seconds.

[0074] Step S2, Multi-dimensional parameter acquisition and preprocessing:

[0075] Data acquisition unit with For a period of time, collect , , , , The parameters are processed sequentially, including moving average filtering, outlier removal, and dimension normalization preprocessing, and the effective parameters are output to the scenario demand index calculation unit.

[0076] Step S3 Scenario Demand Index calculate:

[0077] The scenario demand index calculation unit performs weighted fusion calculations to calculate S in real time based on preprocessed multi-dimensional parameters, and dynamically adjusts the weights according to the real-time power consumption of the chiller and the priority of external tasks. The calculation result S is output to the joint optimization objective calculation unit and the Kalman filter prediction unit.

[0078] Step S4, Kalman filtering and trend prediction:

[0079] Kalman filter prediction unit , , , Perform filtering and noise reduction, complete trend prediction within a 5-minute time window, and output. , , Simultaneously, a temperature transition trajectory is planned (constraining the temperature change rate to ≤2K / min); the correction parameters and temperature transition trajectory are sent back to the scenario demand index calculation unit and the joint optimization target calculation unit, and then synchronously output to the PWM output execution unit.

[0080] Step S5, jointly optimize the target value calculate:

[0081] Joint optimization objective computation unit 43 based on , Construct a dual-objective optimization function and dynamically adjust lifetime weights. With sensitivity weight Calculate the joint optimization objective value The results are then output to the Kalman filter prediction unit and the PWM output execution unit.

[0082] Step S6, Pattern Decision and Parameter Matching:

[0083] Predicted values With threshold , With threshold Comparative judgment: If or The decision-making process is in a high-sensitivity mode, matching the target temperature. PWM duty cycle like and Verification within 5 consecutive minutes In the absence of external high-sensitivity commands, the decision is made to adopt the long-life mode, matching the target temperature. PWM After matching is completed, the mode parameters are output to the PWM output execution unit.

[0084] Step S7: PWM Output and Temperature Closed-Loop Control

[0085] The PWM output execution unit adjusts the PWM duty cycle in steps of ≤5% based on the mode parameters and temperature transition trajectory, generating a control signal output to the refrigerator to control the temperature. Gradually approaching the preset trajectory It achieves closed-loop control with a steady-state accuracy of ±0.3K; the mode switching response time is ≤4s, and the temperature fluctuation during the switching process is ≤0.8K.

[0086] Step S8 (Loop / Termination Check):

[0087] The intelligent control module monitors two types of termination trigger conditions in real time:

[0088] A type of power supply voltage monitoring system determines a power outage when the input voltage is <22V.

[0089] The second type of CAN bus command monitoring is used. When a stop command with command ID=0x789 is received, it is determined to be an active termination. If the termination condition is not triggered, the process returns to step S2 and repeats the execution. If the termination condition is triggered, the PWM output is turned off, the current running data (mode, temperature, power consumption, cumulative running time) is saved to EEPROM, and the temperature control program is terminated.

[0090] In this embodiment, parallel tests were conducted on the intelligent dual-mode solution and the traditional fixed cooling solution (100K constant temperature control, no automatic mode switching function) under the same test environment (25℃, standard atmospheric pressure, no strong electromagnetic interference). The core performance data comparison is summarized in Table 1:

[0091]

[0092] Based on the test data in Table 1, the core technical advantages of the intelligent dual-mode solution in this embodiment compared to the traditional fixed cooling solution are as follows:

[0093] A dynamic balance is achieved between detection performance and equipment lifespan: In high-sensitivity mode, NETD is as low as 11.6mK, which is 37.3% higher than traditional solutions, meeting the requirements of high-precision detection; in long-life mode, the predicted lifespan of the cooler reaches 12,600 hours, which is 110% higher than traditional solutions, solving the industry pain point of "inability to balance sensitivity and lifespan" and adapting to the needs of different mission cycles.

[0094] Significantly optimized temperature control accuracy and switching performance: The steady-state temperature control accuracy of the chip reaches ±0.3K, which is 70% higher than the traditional solution; the mode switching response time is ≤4s, and the temperature fluctuation during the switching process is ≤0.8K with no overshoot oscillation. Thanks to the synergistic effect of Kalman filter trend prediction, temperature transition trajectory planning and PWM duty cycle smooth adjustment (step size ≤5%), imaging stability is ensured.

[0095] Significantly improved energy efficiency and environmental adaptability: In long-life mode, steady-state power consumption is as low as 6.2W, which is 22.5% lower than traditional solutions, demonstrating significant energy efficiency advantages; in high and low temperature environments of -55℃ to 70℃ and 10g random vibration environments, the temperature control deviation is ≤0.2K, and the performance degradation is ≤5% after 1000 hours of long-cycle operation. Relying on hardware modular design and algorithm anti-interference mechanism, it is adapted to complex working conditions.

[0096] High degree of automation and improved ease of operation: Automatic mode switching can be completed without manual intervention. Compared with the manual switching mode of traditional solutions (which takes ≥30 seconds), the response efficiency is greatly improved, meeting the real-time task requirements in dynamic scenarios.

[0097] This embodiment also provides a computer-readable storage medium, specifically including the on-chip Flash (1MB capacity) and external EEPROM (model: AT24C256, 256KB capacity) of the STM32H743VIT6 microprocessor; the storage medium stores a computer program that implements the multi-parameter fusion adaptive predictive control algorithm of the present invention, which, when executed by the microprocessor, can fully implement... Figure 2 The full functionality of the algorithm unit shown, and Figure 3 The intelligent temperature control closed-loop process is shown.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart temperature-controlled cooling infrared detector, characterized in that, It includes a superlattice chip, a refrigerator, a temperature sensing module, an intelligent control module, and a mode switching interface, among which: The superlattice chip is mounted on the cold head of the refrigerator and is used to detect the infrared radiation of the target and output the target signal. The refrigeration unit has its PWM input terminal directly connected to the intelligent control module and is used to cool the Type II superlattice chip. The temperature sensing module outputs a signal that is connected to the intelligent control module to collect temperature information from the Type II superlattice chip and the cold head of the refrigerator. The mode switching interface is connected to the intelligent control module via a communication bus, and is used to receive external task commands and transmit them to the intelligent control module. The intelligent control module, built on a microprocessor, receives various parameter information from the temperature sensing module and the mode switching interface, executes a multi-parameter fusion adaptive predictive control algorithm, and outputs a PWM control signal to control the operating temperature of the refrigerator. The intelligent control module has at least two operating modes: a long-life mode and a high-sensitivity mode.

2. The intelligent temperature-controlled cooling infrared detector according to claim 1, characterized in that, The superlattice chip is a type II superlattice focal plane array chip, and the refrigerator is an integrated Stirling refrigerator.

3. The intelligent temperature-controlled cooling infrared detector according to claim 1 or 2, characterized in that, The intelligent control module includes a data acquisition unit, which is used to collect chip temperature, temperature change rate, target signal strength, external command frequency, and historical task information from the temperature sensing module and mode switching interface.

4. The intelligent temperature-controlled cooling infrared detector according to claim 3, characterized in that, The execution of the multi-parameter fusion adaptive predictive control algorithm includes the following operations: The scenario demand index is calculated by weighting and fusing various parameters based on the temperature change rate, target signal strength, external command frequency, and historical mission change trends to obtain a scenario demand index that represents the comprehensive requirements of the current mission for detection sensitivity and refrigerator lifespan. The joint optimization target calculation combines the predicted lifespan of the chiller with the estimated detection sensitivity, assigns adjustable lifespan weights and sensitivity weights respectively, and performs weighted calculations to obtain the joint optimization target value. The current target working mode is then determined based on the optimization target value. Kalman filtering prediction filters and predicts the trend of the scenario demand index and optimization target value within a preset time window, and determines the temperature transition trajectory and mode switching time based on the prediction results.

5. The intelligent temperature-controlled cooling infrared detector according to claim 4, characterized in that, The adjustment of lifetime weight and sensitivity weight in the joint optimization target calculation follows these rules: when the remaining task duration exceeds a preset duration threshold, the lifetime weight is increased to maintain the refrigerator at a higher operating temperature; when the external command frequency exceeds a preset frequency threshold or the target signal strength is lower than a preset strength threshold, the sensitivity weight is increased to enable the refrigerator to operate at a lower operating temperature.

6. The intelligent temperature-controlled cooling infrared detector according to claim 5, characterized in that, The intelligent control module performs control through the following steps: Power on and initialize, then enter the default long life mode, and set the corresponding target operating temperature range and control parameters of the PWM control signal. Periodically collect chip temperature, temperature change rate, target signal strength, external command frequency, and historical task trend parameters; Determine the current target working mode and target temperature adjustment strategy based on the scenario demand index and joint optimization target value; Based on the prediction results of the Kalman filter, the control parameters of the PWM control signal are adjusted so that the operating temperature of the refrigerator gradually approaches the target temperature along the preset transition trajectory. Repeat the above steps of data collection, calculation, prediction, and adjustment until the system is powered off or a stop command is received.

7. The intelligent temperature-controlled cooling infrared detector according to claim 6, characterized in that, The intelligent control module receives task instructions sent by external devices through a mode switching interface. The task instructions include task type, task priority, and task duration. The intelligent control module adjusts the weight parameters of the scenario demand index and the lifetime and sensitivity weights of the optimization target according to the task instructions, so as to realize the automatic switching between long lifetime mode and high sensitivity mode.

8. The intelligent temperature-controlled cooling infrared detector according to claim 7, characterized in that, The intelligent control module predicts that the target signal strength will drop to a preset threshold within a preset time window based on the Kalman filter, and switches from the long life mode to the high sensitivity mode in advance, so that the refrigerator completes the temperature transition before the target signal strength actually drops.

9. The intelligent temperature-controlled cooling infrared detector according to claim 8, characterized in that, The intelligent control module controls the rate of change of the refrigerator's operating temperature based on the temperature transition trajectory predicted by the Kalman filter, so that the operating temperature gradually approaches the target temperature along the temperature transition trajectory, thereby suppressing temperature fluctuations during mode switching.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it is used to implement the intelligent temperature control function of the intelligent temperature-controlled cooling infrared detector according to any one of claims 1 to 9.

Images