Intelligent control method and system for supercooling of heat pump based on multi-sensor data fusion

Abstract

This invention proposes a smart control method and system for heat pump subcooling based on multi-sensor data fusion. The method includes: synchronously collecting condenser outlet temperature, subcooler inlet and outlet liquid pipe temperatures, compressor discharge pressure, and ambient temperature; estimating and replacing abnormal data based on thermodynamic coupling relationships to form a stable state input; extracting the subcooling temperature difference in the condensing section and the subcooler temperature difference, and constructing a composite risk criterion by combining the discharge pressure and ambient temperature; dynamically generating fan start / stop signals and continuous fan speed commands based on the risk criterion; and using changes in liquid pipe temperature difference during the execution phase to perform soft verification of the actual cooling effect of the fan, achieving real-time correction of the control commands. This invention constructs a closed-loop control structure of perception-judgment-decision-verification-correction, improving the operational stability, safety, and energy efficiency of the heat pump system under complex operating conditions without adding additional sensors.

Description

Technical Field

[0001] This invention belongs to the field of heat pump technology, and particularly relates to a heat pump subcooling intelligent control method and system based on multi-sensor data fusion. Background Technology

[0002] Current air-cooled heat pump systems still suffer from significant shortcomings in heating performance and operational stability under varying environmental conditions. Particularly in complex operating conditions such as low temperatures, high outlet water temperatures, and decreased heat exchanger efficiency, traditional subcooling control methods often fail to guarantee system stability and energy efficiency under dynamic loads. Existing heat pumps generally rely on natural subcooling of the condenser or fixed logic threshold control. These methods are simplistic and have delayed responses. When changes in operating conditions lead to increased condensing pressure or decreased refrigerant flow, the controller cannot intervene promptly, easily resulting in faults such as excessively high compressor discharge temperature, failure of vapor injection enthalpy boosting for liquid extraction, and high-pressure protection failures. Furthermore, because heat pump systems involve multiple signal sources such as condensing temperature, liquid line temperature, discharge pressure, and ambient temperature, sensor drift, asynchronous sampling, and partial failures can all cause control misjudgments, rendering traditional rule-based algorithms unreliable. Especially in actual operation, data from different sensors often exhibit non-linear coupling relationships; for example, condensing temperature and discharge pressure are not in real-time correspondences, and changes in liquid line temperature difference also have time delays. Therefore, control strategies based on a single threshold are insufficient to accurately assess system risk states. Furthermore, existing heat pump controls are mostly open-loop or weakly closed-loop, lacking real-time evaluation of control performance. This leads to frequent start-stop cycles, over-response, or lag in the operation of air-cooled subcoolers, reducing energy efficiency and exacerbating equipment wear. In summary, the key shortcomings of current heat pump subcooling control are: a lack of multi-source information fusion processing capabilities, a lack of dynamic identification of thermodynamic state characteristics, a lack of risk-based adaptive control decision-making mechanisms, and a lack of dynamic correction capabilities for performance. Summary of the Invention

[0003] The purpose of this invention is to design a heat pump subcooling intelligent control method and system based on multi-sensor data fusion, which can realize the unification of multi-source data fusion, risk self-identification, strategy self-optimization and execution self-correction of heat pump systems.

[0004] To achieve the above objectives, a first aspect of the present invention provides a smart control method for heat pump subcooling based on multi-sensor data fusion, the method comprising:

[0005] Collect the condenser outlet temperature, subcooler front liquid line temperature, subcooler rear liquid line temperature, compressor discharge pressure, and ambient temperature in the heat pump system.

[0006] The collected data is subjected to anomaly detection, and the data identified as abnormal is estimated and replaced based on the internal physical coupling relationship of the heat pump system to form an estimated state vector;

[0007] Based on the estimated state vector, the subcooling temperature difference in the condensing section and the subcooler temperature difference are extracted as key feature variables, and a composite risk criterion is constructed by combining the compressor discharge pressure and the ambient temperature.

[0008] Generate fan start / stop signals and target wind speed commands based on the composite risk criteria;

[0009] The fan start / stop signal and the target wind speed command are output to the air-cooled subcooler for execution.

[0010] During execution, the actual operating effect of the fan is verified based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and the target wind speed command for the next cycle is corrected based on the verification results.

[0011] Furthermore, the anomaly detection employs a sliding window mechanism, which calculates the median of the time series formed by each sampled variable and its values ​​in the past two sampling periods, and determines whether the current value deviates from the median by more than a preset threshold.

[0012] Furthermore, when estimating and replacing outlier data, an offline-fitted linear regression model is used, with other physical quantities not marked as outliers as input variables, to calculate the estimated value of the replaced variable.

[0013] Furthermore, the composite risk criterion includes information on compressor discharge pressure, condenser subcooling temperature difference, subcooler temperature difference, and relative deviation between liquid pipe temperature and ambient temperature, and improves the sensitivity of identifying subcooler heat exchange failure or external thermal disturbance conditions through nonlinear enhancement terms.

[0014] Furthermore, the generation of the fan start / stop signal is based on a dynamic response factor, which adaptively adjusts the fan start threshold according to the subcooling temperature difference in the condensing section and the subcooler temperature difference.

[0015] Furthermore, the generation of the target wind speed command simultaneously considers the magnitude of the composite risk criterion, the rate of change of the supercooler liquid pipe temperature difference, and the influence of ambient temperature on the wind speed reference.

[0016] Furthermore, the fan start / stop signal controls the fan drive relay through a digital I / O port, and the target wind speed command controls the fan speed control module through a PWM signal or an analog voltage signal.

[0017] Furthermore, the actual operating effect of the fan is verified by comparing the rate of change of the temperature difference in the subcooler liquid pipe between the current cycle and the previous cycle. If the temperature difference does not increase significantly with the increase of the fan speed command, it is determined that the fan is not operating effectively.

[0018] Furthermore, the correction of the target wind speed command is based on the degree of deviation between the actual operation of the fan and the expected effect. The greater the deviation, the greater the correction. The smoothness of the correction process is controlled by an exponential function.

[0019] A second aspect of the present invention provides a heat pump subcooling intelligent control system based on multi-sensor data fusion, the system comprising:

[0020] The data acquisition module is used to collect the condenser outlet temperature, subcooler front liquid pipe temperature, subcooler rear liquid pipe temperature, compressor discharge pressure, and ambient temperature in the heat pump system; it performs anomaly detection on the collected data, and replaces the data that is determined to be abnormal with an estimated state vector based on the internal physical coupling relationship of the heat pump system.

[0021] The risk assessment module is used to extract the subcooling temperature difference of the condensing section and the subcooler temperature difference as key feature variables based on the estimated state vector, and to construct a composite risk criterion by combining the compressor discharge pressure and the ambient temperature.

[0022] The control decision module is used to generate a fan start / stop signal and a target wind speed command based on the composite risk criteria; and output the fan start / stop signal and the target wind speed command to the air-cooled subcooler for execution.

[0023] The execution verification module is used to verify the actual operating effect of the fan based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and to correct the target wind speed command for the next cycle based on the verification results.

[0024] The beneficial technical effects of the present invention are at least as follows:

[0025] To address the aforementioned issues, this invention provides a heat pump subcooling intelligent control method and system based on multi-sensor data fusion. It utilizes real-time acquisition of data from multiple sensors, including condenser outlet temperature, liquid pipe temperature, exhaust pressure, and ambient temperature. A valuation mechanism based on physical coupling is employed to achieve data fusion and anomaly compensation, forming a highly stable state input vector. Based on this, the system extracts key characteristic variables such as the condenser temperature difference and subcooler temperature difference, and constructs a composite risk criterion by combining exhaust pressure and ambient temperature. Exponential enhancement terms and nonlinear regularization terms are used to accurately identify complex operating conditions such as high pressure, subcooling failure, and external thermal disturbances. Based on the risk criterion results, the system introduces dynamic response factors and risk-driven functions to generate control commands, continuously adjusting the start / stop and speed of the air-cooled fan, thus transitioning from static logic control to dynamic continuous control. During the control command execution phase, the system utilizes the dynamic response of the liquid pipe temperature difference to establish a thermodynamic soft verification mechanism, monitoring and correcting the fan's performance in real time. This achieves self-correction and fault-tolerant control at the execution layer without adding additional sensors. Through the above-mentioned innovative structure, this invention achieves the unification of multi-source data fusion, risk self-identification, strategy self-optimization and execution self-correction in heat pump systems, forming an intelligent subcooling control scheme with high response speed, high robustness and high energy efficiency, which significantly improves the operational stability and control accuracy of heat pump systems in complex environments. Attached Figure Description

[0026] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0027] Figure 1 This is a flowchart of the intelligent control method for heat pump subcooling based on multi-sensor data fusion according to the present invention.

[0028] Figure 2 This is a framework diagram of the intelligent control system for heat pump subcooling based on multi-sensor data fusion according to the present invention. Detailed Implementation

[0029] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0030] In one or more embodiments, such as Figure 1 As shown, a smart control method for heat pump subcooling based on multi-sensor data fusion is disclosed, the method comprising the following:

[0031] S1: Collect the condenser outlet temperature, subcooler front liquid pipe temperature, subcooler rear liquid pipe temperature, compressor discharge pressure, and ambient temperature in the heat pump system; perform anomaly detection on the collected data, and replace the data that is judged to be abnormal with an estimated state vector based on the internal physical coupling relationship of the heat pump system.

[0032] Specifically, the core task of this step is to construct a stable and complete state input vector by synchronously acquiring, structuring, and performing fault-tolerant estimation of multiple key physical quantities during the actual operation of the heat pump system. This provides a unified input basis for subsequent operating condition judgment and intelligent control strategy generation. Due to the complex operating environment of heat pump systems, various sensors are easily affected by factors such as water vapor, frost, and electrical interference, causing data drift, sudden changes, or transmission failures. Therefore, this step must be designed with strong anomaly tolerance and the edge computing resources must be lightweight enough to reliably operate on typical unit-level controllers (such as STM32F4 series, Freescale S12 series, or small embedded platforms based on ARM Cortex-M4 / M7).

[0033] This step processes data from five key sensors, which collect data on the system's heat exchange status, compressor workload, and external thermal environment. The sampling period is uniformly 1 second. The data acquisition interfaces and physical deployment methods of all sensors are as follows: condenser outlet temperature... Data is collected using an NTC thermistor mounted on the condenser's liquid outlet copper tube. The sensor uses a standard 10KΩ resistance value, and the thermistor is bonded to the copper tube with thermally conductive adhesive. It is connected to the analog-to-digital converter interface of the main control board. Liquid tube temperature... and Located at the front and rear ends of the air-cooled subcooler, respectively, both employ waterproof and corrosion-resistant NTC probes attached to the copper tube wall, encapsulated in a plastic shell with a metal base plate to enhance response speed. Sampling is conducted via a voltage divider circuit connected to the ADC channel. Exhaust pressure The pressure is collected by an analog voltage-type pressure transmitter connected to the compressor exhaust port. The output voltage is 0.5~4.5V, corresponding to the actual pressure range. This voltage is then converted into a digital pressure value by the 12-bit ADC module of the electronic control system. Ambient temperature... Installed near the air inlet on the unit casing, it uses an externally encapsulated NTC sensor to reduce interference from sunlight radiation, and its signal is also received through the ADC channel.

[0034] After the above five types of observation data are sampled synchronously, a structured observation vector is formed in the controller's memory:

[0035] ;

[0036] But at any time Some of the five variables mentioned above may exhibit anomalies, such as... Mutations occur The signal sample value deviates from the historical average by more than a set proportion. The system first performs an anomaly detection mechanism based on a sliding window for each variable. For example, reading its current value and the values ​​of the past two periods constitutes a time series. Calculate the median It then checks whether the current value deviates from the median by more than a set threshold (e.g., ±5.0). If so, it is marked as an anomaly. This operation is performed independently for each type of variable, and the result forms a Boolean array of anomaly markers.

[0037] For all variables marked as anomalous, the system does not directly discard their data, nor does it use linear interpolation or retain the previous period's value. Instead, it constructs a lightweight estimation function based on the coupling relationships between physical variables within the heat pump system. This estimation function has a fixed structure and is based on offline fitting of historical data, enabling rapid inference locally within the system. Its basic structure is as follows:

[0038] ;

[0039] in, Indicates the abnormal variable The estimated value, and For other observables that have a stable thermo-mechanical coupling relationship with it under actual operating conditions, These are the linear regression coefficients obtained through offline fitting. For example, when the compressor discharge pressure sensor malfunctions, the system can detect the condenser outlet temperature. and the temperature at the front end of the liquid tube The valuation input is obtained based on the following model:

[0040] ;

[0041] If the current measurement is , Then it can be calculated This is used to replace the current outlier. This replacement logic is triggered by an anomaly flag, performs evaluation only on variables where a problem is detected, and retains non-outlier variables as is. Finally, the values ​​are concatenated into an output vector.

[0042] ;

[0043] Each item The acquisition of values ​​follows the above mechanism: if the original value is valid, it is directly retained; if it is detected as an anomaly, it is calculated by the structured estimation function within the system, which has high stability and consistency.

[0044] S2: Based on the estimated state vector, extract the subcooling temperature difference of the condensing section and the subcooler temperature difference as key feature variables, and construct a composite risk criterion by combining the compressor discharge pressure and the ambient temperature;

[0045] Specifically, this step follows the process of constructing the estimated state vector in the previous step. The goal is to extract key features from these five types of physical quantities that can comprehensively describe the current thermodynamic state and potential failure risks of the heat pump system without adding additional sensors, and to construct an operational risk criterion tightly coupled to the subcooling control objective based on these features. This step is the core of the entire control process, enabling "intelligent judgment on whether active subcooling adjustment is needed." Its design determines whether the controller has the ability to accurately identify control requirements under different operating scenarios, such as whether it can intervene in advance to adjust under conditions of high compressor load, condenser scaling, or subcooler failure, thereby effectively avoiding high-pressure failures, inefficient operation, or system protection shutdowns.

[0046] Unlike general HVAC condition assessments, heat pump subcooling control exhibits two unique data patterns: first, "strong coupling but significant hysteresis between thermodynamic parameters," for example, the relationship between condensing temperature and exhaust pressure is not instantaneous and linear; second, "the determination of local subcooling failure must rely on temperature difference rather than absolute value," such as under high-temperature conditions. While the temperature is high, it may still be effectively supercooled. Therefore, using only the original physical quantities or their differences is insufficient to accurately identify the risk state of the system. This step proposes a feature combination structure that combines physical laws, structural characteristics, and actual control objectives. By constructing a composite quantity with practical significance and introducing a nonlinear enhancement term with physical interpretation capabilities into the risk index, the sensitivity and reliability of risk identification are improved.

[0047] First, at the level of basic feature construction, two key temperature difference features are extracted from the state vector: the subcooling temperature difference in the condensation section. Temperature difference with subcooler :

[0048] ;

[0049] in This represents the difference between the condenser outlet temperature and the liquid pipe temperature at the front of the subcooler, used to determine whether the condenser has formed an effective natural subcooling section. This represents the temperature difference between the inlet and outlet liquid pipes of the air-cooled subcooler, used to determine whether the subcooler currently possesses effective heat exchange performance. A value that is too small (e.g., less than 4) indicates a weak condensation process, while a value that is too small (e.g., less than 3) indicates a decrease in the air-cooled heat exchange efficiency of the subcooler. All temperature data are derived from the estimated state vector of the previous stage, without the addition of additional sensors.

[0050] To more effectively integrate the aforementioned temperature difference information with the system operating load, this step proposes a composite risk criterion structure. This structure introduces a liquid-pipe heat transfer penalty term and a nonlinear regularization term on the basis of the traditional pressure-temperature difference normalization index, thereby enabling the controller to make more accurate adjustment judgments at the edge of the operating condition.

[0051] ;

[0052] in: The compressor discharge pressure is derived from the estimated state vector. This refers to the heat exchange capacity of the condenser; if it is too small, the risk increases. To prevent the denominator from approaching zero, the value of the differential constant is taken as follows: ; This is a heat transfer penalty term that increases rapidly as the temperature difference of the subcooler decreases, used to enhance the system's response sensitivity in the event of subcooler failure. This is used to measure the relative deviation between the liquid pipe temperature and the ambient temperature, reflecting whether the current heat exchange of the system is severely limited by external operating conditions; and The weighting parameters set for experience are suggested to have values ​​such as... , , To prevent division by zero, the value of the micro constant is as follows: .

[0053] This criterion structure introduces two physically reasonable enhancements on the basis of the traditional "pressure divided by temperature difference" form: (1) This causes the system risk index to rise exponentially when the subcooler fails (temperature difference approaches 0), thereby triggering forced subcooling control; (2) the normalized deviation term of the liquid pipe and ambient temperature reflects the special risk scenario of "high external temperature and low subcooling", which is a unique identification indicator of "external environmental interference dominating internal heat exchange failure" in heat pump systems. This combination allows the controller to accurately identify multiple "atypical high-risk states" without introducing external models or complex calculation logic, avoiding misjudgments caused by relying solely on pressure or absolute temperature values. Finally, this step outputs two variables: feature vector Composite risk criteria Both serve as inputs for control strategy matching and decision generation in the next step.

[0054] S3: Generate a fan start / stop signal and a target wind speed command based on the composite risk criterion; output the fan start / stop signal and the target wind speed command to the air-cooled subcooler for execution;

[0055] Specifically, this step builds upon the operational features established in the previous stage. Risk criteria Based on this, control commands for driving the air-cooled subcooler fan are generated. This step is the most crucial decision-making node in the entire intelligent subcooling control system for heat pumps. It determines whether the system needs active subcooling adjustment and, under what risk level, what fan operation strategy should be adopted to balance energy efficiency and safety. In the air-cooled subcooler control scenario addressed by this solution, conventional control logic, primarily based on "fixed temperature difference threshold triggering + segmented fan speed adjustment," struggles to adapt to rapid changes in system load, limited condensing efficiency, and strategy lag issues caused by sensor drift. This step constructs a fusion-criteria-driven continuous controller based on the risk response curve, ensuring both decisive start-up and shutdown actions and achieving continuous and dynamic robustness in fan speed adjustment.

[0056] The control logic first uses the risk criterion output from the previous stage. As a global driving force, combined with and Two key temperature difference variables are used to construct control start / stop determination and airflow adjustment curves, respectively. Unlike traditional fixed threshold control methods, this step no longer uses a single numerical threshold for fan start / stop determination, but instead introduces a dynamic response factor based on the current environment and condensation heat transfer level. By constructing a nonlinear mapping function to adaptively adjust the fan start point, the controller can exhibit differentiated response capabilities under varying ambient temperatures and condensing capacities. This mapping function is as follows:

[0057] ;

[0058] in Control the steepness of the response curve (e.g., 1.5~2.0). This represents the design value for the target condensation temperature difference (e.g., 7.0). This is a fault sensitization factor for the air-cooled heat exchanger (e.g., 1.0~3.0). The core idea behind this function design is: when the condensing temperature difference is too small or the subcooler temperature difference is abnormal, As the value increases, the fan is more easily triggered. The entire logic strikes a balance between control stability and control sensitivity.

[0059] In fan start / stop determination and By combining the driving variables for control threshold matching, the following start / stop functions are constructed:

[0060] ;

[0061] in This indicates the fan start / stop signal (1 for on, 0 for off). With a fixed control threshold (e.g., 6.5), all variables are calculated at the current moment, ensuring the controller can respond promptly based on the fusion status in each sampling cycle. Compared to traditional methods... The fixed threshold control method integrates the heat exchange efficiency of the air cooler, the condensing capacity, and the system risk load, enabling the fan to start in advance before the risk has fully accumulated, while delaying the start-up in high condensing capacity scenarios to reduce energy consumption.

[0062] Furthermore, while the fan is running, the controller needs to output continuous fan speed commands to dynamically adjust the airflow. This step does not employ multi-segment logic; instead, it constructs a hybrid control structure based on the risk value and the rate of change of the liquid pipe temperature difference. Its control logic is as follows:

[0063] ;

[0064] in: This is a risk-driving factor (e.g., 10.0), used to amplify risk. Linear response to wind speed; The dynamic cooling rate feedback suppression coefficient (e.g., 30.0) is used when... Reduce wind speed if the descent is too rapid; Control the linkage compensation coefficient between wind speed and ambient temperature (e.g., 5.0~10.0). The standard reference ambient temperature is (e.g., 30.0).

[0065] The wind speed regulation here is not only based on risk criteria, but also incorporates the derivative of the liquid pipe temperature difference change as a reverse braking mechanism for the system's active response capability. Simultaneously, the ambient temperature is mapped using a sine function to adjust the wind speed baseline, automatically raising the minimum wind speed under high ambient temperatures to enhance adaptability to external heat loads. This structure ensures that the wind speed increase rate is reduced when the system's subcooling effect has begun to manifest, preventing over-adjustment; it also guarantees a minimum wind speed guarantee capability under high-temperature conditions, avoiding compressor protection shutdowns caused by environmental thermal barriers.

[0066] The final control command consists of two quantities: a fan start / stop signal and a fan stop signal. With target air volume The system outputs on the local controller in the following structure:

[0067] ;

[0068] Control commands are sent to the air cooler execution layer via the PWM speed control port or the fan speed control drive module to complete the real-time adjustment of the actual air volume.

[0069] S4: During execution, the actual operating effect of the fan is verified based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and the target wind speed command for the next cycle is corrected based on the verification results.

[0070] Specifically, this step is based on the control commands generated in the previous stage. This process physically executes the control intent and establishes a closed-loop control behavior structure with response confirmation, dynamic correction, and fault tolerance. Unlike traditional control methods that directly convert control commands into hardware level signals or PWM signals, this step not only focuses on whether the control command has been "sent out," but also on whether the fan "actually moves" and "moves correctly." If this step fails, even with the most accurate previous judgments, the system cannot truly achieve overcooling regulation control. This step introduces a "soft verification" mechanism based on heat exchange reaction, using existing temperature difference measurement points within the system to indirectly determine whether the fan is actually running and producing an effect. This avoids the need for additional hardware such as wind speed sensors, improving deployability and system cost control.

[0071] In practical operation, the system follows the control commands. The value controls the fan's start / stop status. This signal is output as a high or low level through a digital I / O port in the edge controller, controlling the on / off state of the fan drive relay or fan speed regulator. (Fan speed regulation value) The PWM signal or 0–10V analog signal is then output to the fan speed control board to control the actual percentage of the fan's operating speed. The duty cycle of the PWM signal and The values ​​are linearly related, for example At that time, a PWM waveform with a 60% duty cycle is output. This signal can be output by the hardware PWM module of a microcontroller such as STM32, or generated by a DAC chip or a PWM-to-analog signal conversion module.

[0072] However, in actual operation, fan response has a certain degree of electrical inertia and mechanical lag. For example, in high-temperature environments during summer, voltage fluctuations and fan motor temperature rise protection may cause the PWM signal to be issued, but the fan may not start immediately, or even fail to start at all. In this situation, if the control system does not have an execution confirmation mechanism, the system will misjudge that the subcooler has started, thus entering a risky area.

[0073] To avoid such miscontrol, this step incorporates a "fan execution effect confirmation mechanism". This mechanism does not rely on a dedicated fan speed sensor. Instead, it utilizes the existing liquid pipe temperature sensor in the system to determine whether the fan is producing the desired cooling effect by dynamically changing the temperature difference between the liquid pipes before and after the subcooler. Specifically, the system records the liquid pipe temperature difference between the current cycle and the previous cycle. and Calculate its rate of change and normalize it:

[0074] ;

[0075] in The temperature difference between the inlet and outlet liquid pipes of the air cooler, as defined in step two, is collected from the NTC temperature sensor already installed in the unit and read in real time through an analog-to-digital converter module (such as the ADC of STM32). The target wind speed command is calculated by the control strategy module from the previous step. To prevent the use of tiny constants with a denominator of zero, a value such as 1.0 is taken. If... A significantly negative value indicates that the temperature difference did not increase with the increase in wind speed, possibly because the fan was not working or the wind speed was not up to standard; if A value close to zero indicates limited fan effectiveness; a positive value with a good growth rate indicates the fan is operating normally. This value is recorded in the controller cache and participates in subsequent execution corrections.

[0076] When the actual operation of the fan deviates from the control intention, the system needs to automatically correct it in the next control cycle. Therefore, this step introduces a correction function. ,exist Based on this, make dynamic adjustments:

[0077] ;

[0078] This structure implements a wind speed self-adjustment mechanism based on the degree of response. This indicates the maximum adjustable wind speed range, such as 20. This is for positive and negative feedback directional control; the exponential term ensures a smooth system response when the deviation is small and rapid correction when the deviation is large, avoiding over-adjustment that could cause wind speed oscillations. For example, when... ,but ,but This indicates that the system will reduce the fan speed in the next cycle and attempt to restore the heat exchange capacity of the air cooler by increasing the liquid pipe temperature difference. All correction logic is executed locally on the controller, requiring no network connection or external intervention, and has instant response capability.

[0079] Ultimately, the actual control commands issued by the system in each control cycle are: This instruction not only retains the regulatory intent of the previous step, but also integrates the hot response confirmation of the actual execution status in this step, thus upgrading the control loop from "judgment-control" to a full-process closed-loop mechanism of "judgment-control-verification-correction".

[0080] In one or more embodiments, such as Figure 2 As shown, a heat pump subcooling intelligent control system based on multi-sensor data fusion is disclosed, the system comprising:

[0081] The data acquisition module is used to collect the condenser outlet temperature, subcooler front liquid pipe temperature, subcooler rear liquid pipe temperature, compressor discharge pressure, and ambient temperature in the heat pump system; it performs anomaly detection on the collected data, and replaces the data that is determined to be abnormal with an estimated state vector based on the internal physical coupling relationship of the heat pump system.

[0082] The risk assessment module is used to extract the subcooling temperature difference of the condensing section and the subcooler temperature difference as key feature variables based on the estimated state vector, and to construct a composite risk criterion by combining the compressor discharge pressure and the ambient temperature.

[0083] The control decision module is used to generate a fan start / stop signal and a target wind speed command based on the composite risk criteria; and output the fan start / stop signal and the target wind speed command to the air-cooled subcooler for execution.

[0084] The execution verification module is used to verify the actual operating effect of the fan based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and to correct the target wind speed command for the next cycle based on the verification results.

[0085] It is worth noting that the specific workflow of the intelligent control system for heat pump subcooling based on multi-sensor data fusion provided in this embodiment of the invention is the same as that of the intelligent control method for heat pump subcooling based on multi-sensor data fusion described in the above embodiments, and will not be repeated here.

[0086] This invention also provides a heat pump subcooling intelligent control device based on multi-sensor data fusion, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiments of the heat pump subcooling intelligent control method based on multi-sensor data fusion, for example... Figure 1 The steps S1 to S4 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0087] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the heat pump subcooling intelligent control device based on multi-sensor data fusion.

[0088] The intelligent control device for heat pump subcooling based on multi-sensor data fusion can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. This intelligent control device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the intelligent control device may also include input / output devices, network access devices, buses, etc.

[0089] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASACs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the heat pump subcooling intelligent control device based on multi-sensor data fusion, connecting all parts of the device via various interfaces and lines.

[0090] The memory can be used to store the computer program and / or modules. The processor implements various functions of the heat pump subcooling intelligent control device based on multi-sensor data fusion by running or executing the computer program and / or modules stored in the memory and calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created based on the operation of the air conditioner controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0091] The integrated module of the heat pump subcooling intelligent control device based on multi-sensor data fusion, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0092] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0093] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A heat pump subcooling intelligent control method based on multi-sensor data fusion, characterized in that, The method includes: Collect the condenser outlet temperature, subcooler front liquid line temperature, subcooler rear liquid line temperature, compressor discharge pressure, and ambient temperature in the heat pump system. Anomaly detection is performed on the collected data, and the data identified as abnormal is estimated and replaced based on the internal physical coupling relationship of the heat pump system. An estimated state vector is formed based on the estimated data and the data that were not detected as abnormal. Based on the estimated state vector, the subcooling temperature difference in the condensing section and the subcooler temperature difference are extracted as key feature variables, and a composite risk criterion is constructed by combining the compressor discharge pressure and ambient temperature from the estimated state vector. The composite risk criterion includes the compressor discharge pressure, the subcooling temperature difference in the condensing section, the subcooler temperature difference, and the relative deviation information between the liquid pipe temperature at the front end of the subcooler and the ambient temperature from the estimated state vector. The sensitivity of identifying subcooler heat exchange failure or external thermal disturbance conditions is improved by using a nonlinear enhancement term. Fan start / stop signals are generated based on the composite risk criterion and dynamic response factor, wherein the dynamic response factor is a nonlinear mapping function constructed based on the subcooling temperature difference of the condensing section and the subcooler temperature difference. The target wind speed command is generated based on the magnitude of the composite risk criterion, the rate of change of the temperature difference in the supercooler liquid tube, and the influence of ambient temperature on the wind speed reference. The fan start / stop signal and the target wind speed command are output to the air-cooled subcooler for execution. During execution, the actual operating effect of the fan is verified based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and the target wind speed command for the next cycle is corrected based on the verification results.

2. The intelligent control method for heat pump subcooling based on multi-sensor data fusion according to claim 1, characterized in that, The anomaly detection employs a sliding window mechanism, which calculates the median of the time series formed by each sampled variable and its values ​​in the past two sampling periods, and determines whether the current value deviates from the median by more than a preset threshold.

3. The intelligent control method for heat pump subcooling based on multi-sensor data fusion according to claim 1, characterized in that, When estimating and replacing outlier data, an offline-fitted linear regression model is used, with other physical quantities not marked as outliers as input variables, to calculate the estimated value of the replaced variable.

4. The intelligent control method for heat pump subcooling based on multi-sensor data fusion according to claim 1, characterized in that, The fan start / stop signal controls the fan drive relay through a digital I / O port, and the target wind speed command controls the fan speed control module through a PWM signal or an analog voltage signal.

5. The intelligent control method for heat pump subcooling based on multi-sensor data fusion according to claim 1, characterized in that, The actual operating effect of the fan is verified by comparing the rate of change of the temperature difference in the subcooler liquid pipe between the current cycle and the previous cycle. If the temperature difference does not increase significantly with the increase of the fan speed command, the fan is determined to be not operating effectively.

6. The intelligent control method for heat pump subcooling based on multi-sensor data fusion according to claim 1, characterized in that, The correction of the target wind speed command is based on the degree of deviation between the actual operation of the fan and the expected effect. The greater the deviation, the greater the correction. The smoothness of the correction process is controlled by an exponential function.

7. A heat pump subcooling intelligent control system based on multi-sensor data fusion, characterized in that, The system includes: The data acquisition module is used to collect the condenser outlet temperature, subcooler front liquid pipe temperature, subcooler rear liquid pipe temperature, compressor discharge pressure, and ambient temperature in the heat pump system; it performs anomaly detection on the collected data, and performs estimated replacement on the data identified as abnormal based on the internal physical coupling relationship of the heat pump system, and forms an estimated state vector based on the estimated data and the data that was not detected as abnormal. The risk assessment module is used to extract the subcooling temperature difference of the condensing section and the temperature difference of the subcooler as key feature variables based on the estimated state vector, and to construct a composite risk criterion by combining the compressor discharge pressure and ambient temperature from the estimated state vector. The composite risk criterion includes the compressor discharge pressure, the subcooling temperature difference of the condensing section, the temperature difference of the subcooler, and the relative deviation information between the liquid pipe temperature at the front end of the subcooler and the ambient temperature from the estimated state vector. The module also improves the sensitivity of identifying subcooler heat exchange failure or external thermal disturbance conditions through a nonlinear enhancement term. The control decision module is used to generate fan start / stop signals based on the composite risk criterion and dynamic response factor, wherein the dynamic response factor is a nonlinear mapping function constructed based on the subcooling temperature difference of the condensing section and the temperature difference of the subcooler; generate a target wind speed command based on the magnitude of the composite risk criterion, the rate of change of the subcooler liquid pipe temperature difference, and the influence of ambient temperature on the wind speed reference; and output the fan start / stop signals and the target wind speed command to the air-cooled subcooler for execution. The execution verification module is used to verify the actual operating effect of the fan based on the dynamic change of the temperature difference between the liquid pipes before and after the subcooler, and to correct the target wind speed command for the next cycle based on the verification results.

Images