Heating medium circulation efficient heat exchange system and method based on intelligent regulation and control

Abstract

The invention relates to a heating medium circulation efficient heat exchange system and method based on intelligent regulation and control, and the system comprises a heating medium circulation loop, a heat exchange unit, a multi-parameter monitoring module, an intelligent regulation and control module, an execution mechanism, a heating medium state optimization module and an automatic cleaning device. A waste liquid recovery unit is arranged. Dynamic matching of heating medium circulation and heat exchange requirements is achieved through multi-parameter real-time collection and preprocessing, working condition intelligent recognition, heat exchange load prediction, closed-loop regulation and control and heating medium state optimization, a PID neural network and LSTM hybrid prediction algorithm is adopted, and the problems that regulation and control are lagged and multi-parameter collaboration is poor in a traditional system are solved; through all-state monitoring and optimization of the heating medium and intelligent cleaning linkage of scales on the heat exchange surface, the heat exchange efficiency is remarkably improved, the energy consumption is reduced by%, the service life of the heating medium is prolonged, and the system is suitable for various heat exchange scenes such as industrial production and central heating and has extremely high practicability and popularization value.

Description

Technical Field

[0001] This invention relates to the field of heat exchange technology, specifically to a high-efficiency heat exchange system and method based on intelligent control of heat medium circulation. Background Technology

[0002] Heat exchange systems with circulating heat transfer media are core equipment in industrial production and residential heating, and their heat exchange efficiency directly affects energy utilization and production and operating costs.

[0003] However, traditional systems often employ feedback control based on preset thresholds, focusing only on a single parameter (such as the outlet temperature of the heat transfer medium). This makes them unable to adapt to dynamic changes in heat exchange load, resulting in control lag and fluctuations in heat exchange efficiency. They also fail to consider the comprehensive impact of parameters such as heat transfer medium viscosity, purity, and scale thickness on heat exchange performance, leading to unreasonable parameter matching and high energy consumption. Long-term circulation of the heat transfer medium can easily cause aging and deterioration, resulting in a decline in heat exchange performance. Existing systems lack real-time monitoring and dynamic optimization methods. They often rely on periodic manual cleaning, which not only affects continuous operation but also fails to accurately address the actual scale buildup, leading to a continuous decrease in the heat transfer coefficient. Traditional PID control algorithms struggle to cope with the nonlinear and large lag characteristics under complex operating conditions, resulting in low control accuracy. Therefore, a high-efficiency heat exchange system and method based on intelligent control of heat medium circulation is developed to solve the above problems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a high-efficiency heat exchange system and method for heat medium circulation based on intelligent regulation. It has the advantages of intelligent predictive regulation, multi-parameter collaborative optimization, dynamic control of heat medium status, and intelligent scaling treatment. It solves the problems of lagging regulation, poor multi-parameter coordination, uncontrolled heat medium status, inefficient scaling treatment, and high energy consumption in existing heat medium circulation heat exchange systems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control, comprising: Heat medium circulation loop: used for the circulation and transportation of heat medium, including 304 stainless steel circulation pipes with a volume of 5-10m³. 3 The heat storage tank and one-way valve control the flow rate of the heat medium at 0.5-3 m / s. The heat medium is selected from heat transfer oil, ethylene glycol aqueous solution or molten salt. Heat exchange unit: installed in the heat medium circulation loop, adopting shell and tube type (pressure rating 1.6-4.0MPa) or plate type (heat exchange area 10-100m²). 2 The heat exchanger has a heat medium channel and a heat exchange medium channel arranged in counter-current flow, and both ends are equipped with temperature and pressure detection interfaces. Multi-parameter monitoring module: integrated temperature, pressure, flow, viscosity, purity, moisture content and fouling thickness sensors, data acquisition cycle is 0.1-1s, transmission delay ≤0.3s; Intelligent control module: in communication connection with the multi-parameter monitoring module, built-in working condition recognition model (random forest / SVM classification model) and PID neural network and LSTM mixed prediction control algorithm, supporting offline operation and online update; Actuator: including variable frequency circulating pump (flow 0-100m 3 / h, lift 10-50m), electric three-way regulating valve (adjusting accuracy ±1%), 0-50kW adjustable power heat medium heating unit and air-cooled / water-cooled cooling unit; Heat medium state optimization module: containing 5-10μm precision filter, vacuum dehydrator and automatic metering pump, for heat medium purification and performance repair; Automatic cleaning device: linked with the intelligent control module, using 8-12MPa high-pressure spraying or ultrasonic cleaning method, equipped with waste liquid recovery unit.

[0006] Further, the multi-parameter monitoring module includes PT100 temperature sensor (-50~300℃, accuracy ±0.1℃), diffusion silicon pressure sensor (0-2MPa, accuracy ±0.5%FS), electromagnetic flowmeter (0-100m 3 / h, accuracy ±0.3%), vibrating viscosity sensor (5-1000mPa・s, accuracy ±2%), near-infrared spectral purity sensor (90%-100%, accuracy ±1%) and ultrasonic fouling thickness sensor (0-5mm, accuracy ±0.01mm).

[0007] Further, the intelligent control module uses an industrial-grade PLC controller (such as Siemens S7-1500) or an edge computing gateway, built-in Python running environment and ≥32GB data storage unit, supporting offline operation and algorithm online update, control response time ≤1s.

[0008] Further, the heat medium heating unit is electric heating or steam heating, and the cooling unit is air-cooled or water-cooled heat exchanger; the additive supplementing device uses an automatic metering pump with a range of 0-5L / h and an accuracy of ±0.1L / h, and the additive is a heat conducting oil special antioxidant, anti-fouling agent or molten salt stabilizer.

[0009] Further, the intelligent control module supports Ethernet, ModbusRTU, 4G / 5G and LoRa wireless communication protocols, data transmission uses AES-128 encryption, has breakpoint resume transmission function, ensures parameter transmission safety and continuity.

[0010] Further, the heat medium state optimization module controls through multi-parameter linkage: when the heat medium water content is ≥0.3% and the purity is ≤94%, the dehydration device and the additive supplement program are started simultaneously, and the additive supplement amount is dynamically adjusted according to the real-time purity of the heat medium, and the adjustment precision is ±0.05 L / h.

[0011] An intelligent control heat medium circulation efficient heat exchange method, comprising the following steps: S1: multi-parameter acquisition and pretreatment: acquiring operating parameters through a multi-parameter monitoring module, and transmitting the operating parameters to an intelligent control module after removing abnormal values through a 3σ criterion, smoothing through a moving average method, and normalization processing; S2: intelligent identification of working conditions: calling a trained working condition identification model to identify low load (≤30% rated load), medium load (30%-70%), high load (≥70%), and fluctuating load (≥10% / min change rate), and the identification accuracy is ≥95%; S3: heat exchange load prediction: predicting the load change trend in the next 5-10 minutes based on an LSTM model, and the model mean square error MSE is ≤0.01; S4: generation of control instructions: calculating the optimal heat medium flow, temperature, and circulating pressure parameters through a PID neural network algorithm, and converting them into control instructions of an actuator; S5: closed-loop control and state optimization: after the actuator responds to the instructions, the multi-parameter monitoring module feeds back data in real time, and dynamically corrects the control instructions; when the heat medium state optimization or automatic cleaning conditions are triggered, the corresponding modules are started; S6: continuous optimization: periodically updating the working condition identification model and the prediction algorithm parameters to improve the control precision.

[0012] Further, the PID neural network algorithm includes an input layer of 10 neurons (5 historical load data+5 real-time parameters), 3 hidden layers (16-32 neurons per layer), and an output layer of 3 neurons, is trained using an Adam optimizer, the number of iterations is ≥1000 times, and the gradient descent method is used to optimize the objective function of "maximum heat exchange efficiency+minimum energy consumption".

[0013] Further, in S5, the heat medium state optimization trigger conditions are purity ≤92%, viscosity ≥35 mPa・s, or water content ≥0.5%, the filter working pressure is ≤0.3 MPa during purification, and the vacuum degree of the dehydration device is ≥-0.09 MPa; the automatic cleaning trigger conditions are a scale thickness ≥0.5 mm, a cleaning time of 5-15 min, and a heat medium flow maintained at 70%-80% of the rated value during cleaning.

[0014] Further, the LSTM model in the S3 adopts 3 layers of hidden layers (24-32 neurons per layer), inputting load data, heat medium temperature and heat exchange medium flow parameters of 5 time points, optimizing model parameters through 10-fold cross-validation, directly outputting the prediction result when the prediction error is less than or equal to 5%, and starting PID neural network auxiliary correction when the error is greater than 5%.

[0015] Compared with the prior art, the technical scheme of the application has the following beneficial effects: 1. The application realizes dynamic matching of heat medium circulation and heat exchange demand through multi-parameter real-time acquisition and preprocessing, working condition intelligent identification, heat exchange load prediction, closed-loop regulation and control and heat medium state optimization, solves the problems of traditional system regulation and control lag and poor multi-parameter coordination by using a mixed prediction algorithm of PID neural network and LSTM, significantly improves heat exchange efficiency, reduces energy consumption, prolongs the service life of the heat medium, is suitable for various heat exchange scenes such as industrial production and central heating, and has strong practicality and popularization value.

[0016] 2. The application adopts a predictive regulation algorithm, the regulation response time is less than or equal to 1s, solves the problem of traditional feedback regulation lag, adapts to fluctuating load scenes, improves heat exchange efficiency through multi-parameter coordinated regulation + heat medium state optimization + intelligent scaling, avoids invalid energy consumption through load prediction and accurate regulation, reduces the overall energy consumption of the system, reduces heat medium aging and deterioration through real-time monitoring and dynamic optimization, prolongs the service life and reduces the operation cost, realizes the full-process automation of working condition identification, load prediction, regulation execution and state optimization without manual intervention, reduces the operation and maintenance intensity, can adapt to various heat media such as heat conducting oil and ethylene glycol, is suitable for various scenes such as industrial production, central heating and solar heat utilization, and has strong compatibility. BRIEF DESCRIPTION OF DRAWINGS

[0017] Fig. 1 It is a framework diagram of the heat medium circulation efficient heat exchange system based on intelligent regulation and control of the application; Fig. 2 It is a flowchart of the heat medium circulation efficient heat exchange method based on intelligent regulation and control of the application; Fig. 3 It is an implementation flowchart of the application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0019] Please refer toFigs. 1-3 The heat medium circulation high-efficiency heat exchange system based on intelligent regulation in the embodiment comprises: The heat medium circulation loop is composed of a 304 stainless steel circulation pipeline, a heat storage tank (volume 5-10 m³), a one-way valve and a sealing element, the flow rate of the heat medium is controlled at 0.5-3 m / s, and the uniformity of heat transfer is ensured. The heat exchange unit is selected from a plate heat exchanger (heat exchange area 10-100 m²) or a tube-shell heat exchanger (pressure grade 1.6-4.0 MPa), the heat medium passage and the heat-exchanged medium passage are arranged in counter flow, and the heat transfer coefficient is improved. The multi-parameter monitoring module integrates temperature, pressure, flow, viscosity, purity, moisture content and fouling thickness sensors, and the data is transmitted through 4G / 5G or wired Ethernet with a transmission delay of ≤0.3 s. The intelligent regulation module is composed of an industrial PLC controller and an edge computing module, and is internally provided with a working condition recognition model (a random forest model trained based on 10,000+ groups of historical data) and a PID-LSTM hybrid algorithm, and supports offline operation and online update. The actuator is a variable frequency circulating pump (flow rate 0-100 m³ / h, lift 10-50 m), an electric three-way regulating valve (regulation accuracy ±1%, response time ≤0.5 s), an adjustable power heating unit and a high-efficiency cooler, which realize accurate regulation of the heat medium parameters. The heat medium state optimization module is provided with a 5-10 μm precision filter to remove impurities, a vacuum dehydrator to reduce the moisture content (≤0.3%), and an automatic metering pump to supplement antioxidants / anti-fouling agents, so as to maintain the stability of the heat medium performance. The automatic cleaning device is a high-pressure spraying system (nozzle pressure 8-12 MPa) or an ultrasonic cleaning device, which is provided with a waste liquid recovery unit to avoid secondary pollution. The heat medium circulation high-efficiency heat exchange method based on intelligent regulation comprises the following specific steps: Data acquisition and pretreatment: the sensor collects data every 0.1-1 s, removes abnormal values through the 3σ criterion, smoothes noise by the moving average method (window size 5-10), and normalizes to the [0, 1] interval to ensure data reliability. Intelligent recognition of working conditions: the pretreated data is input into the random forest model, the model is optimized through 10-fold cross-validation, and the working condition type and load grade are output, and the identification time is ≤0.1 s. Load prediction: the LSTM model adopts 3 layers of hidden layers (24-32 neurons per layer), inputs historical 5-time load data and real-time parameters, and outputs future 5-10 min load prediction values, and the prediction error is ≤5%. Regulation instruction generation: PID neural network algorithm combines load prediction value and heat medium state parameters, optimizes objective function (maximize heat exchange efficiency + minimize energy consumption) through gradient descent method, calculates optimal heat medium flow, temperature and pressure parameters, and converts them into control signals recognizable by actuators; Closed-loop regulation and state optimization: after the actuators respond to the instructions, the sensors provide real-time feedback of the operating data, the intelligent control module calculates the actual heat exchange efficiency (K = Q / (A Δt), Q is the heat exchange amount, A is the heat exchange area, and Δt is the logarithmic mean temperature difference), if the efficiency is lower than 90% of the rated value, the regulation parameters are dynamically corrected; when the heat medium purity is ≤92%, the viscosity is ≥35 mPa・s or the water content is ≥0.5%, the heat medium purification and additive supplement program is started; when the fouling thickness is ≥0.5 mm, the automatic cleaning is started, and the heat medium flow is maintained at 70%-80% of the rated value during the cleaning process to ensure the continuity of heat exchange; Continuous optimization: the intelligent control module automatically extracts operating data every day, updates the working condition recognition model and prediction algorithm parameters, iteratively optimizes the control strategy, and improves the long-term operation stability.

[0020] Example 1: Heat conducting oil heat exchange system of chemical enterprise (I) System configuration Heat medium circulation loop: 304 stainless steel pipeline (DN100), heat storage tank volume 5m 3 , heat medium is L-QB300 heat conducting oil; Heat exchange unit: plate heat exchanger (heat exchange area 50 2 , pressure grade 1.6MPa), heat medium channel and chemical reaction kettle discharge channel are arranged in countercurrent; Multi-parameter monitoring module: PT100 temperature sensor (-50~300℃), diffusion silicon pressure sensor (0-2MPa), electromagnetic flowmeter (DN100, 0-100m³ / h), vibration type viscosity sensor (5-1000mPa・s), near-infrared purity sensor (90%-100%), ultrasonic fouling thickness sensor (0-5mm); Intelligent control module: Siemens S7-1500PLC+EdgeComputing module, built-in Python3.8 running environment, deploying random forest working condition recognition model and PID-LSTM hybrid algorithm; Actuator: ISG100-160 variable frequency circulating pump (flow 0-100m³ / h, power 11kW), DN100 electric three-way regulating valve, 50kW adjustable electric heating unit, 20㎡ water-cooled cooler; Heat medium state optimization module: 5μm precision filter, vacuum dehydrator (processing capacity 5m 3 / h), automatic metering pump (0-5L / h), additive is T501 heat conducting oil antioxidant; Automatic cleaning device: 10 MPa high-pressure spray system, 16 nozzles, waste liquid recovery tank volume 1 m 3 . (II) Implementation process Initial conditions: The heat exchange medium is the discharge of a chemical reaction kettle (temperature 80℃, flow rate 30m³ / h), which needs to be heated to 150℃; the heat medium initial temperature is 200℃, flow rate 50m³ / h, pressure 1.2MPa, viscosity 20mPa・s, purity 98%, and scale thickness 0.1mm; Data acquisition and preprocessing: The sensor collects data every 0.5s, and after removing abnormal values by 3σ criterion and smoothing by moving average method, it is normalized and transmitted to the intelligent control module; Condition identification: The model identifies as medium load condition (50% rated load), and the identification time is 0.08s; Load prediction: The LSTM model predicts that the heat exchange medium flow rate will increase to 40m³ / h (load increases to 70%) in the next 8min, with a prediction error of 3.2%; Control instruction generation: The PID neural network algorithm calculates the optimal heat medium flow rate 65m 3 / h, temperature maintains 200℃, and the circulating pressure is 1.3MPa, which sends instructions to the frequency conversion circulating pump (frequency conversion to 35Hz) and the electric three-way regulating valve; Closed-loop control: After execution, the feedback data shows that the outlet temperature of the heat exchange medium stabilizes at 150℃, and the heat exchange efficiency increases from 85% to 92%; State optimization: After running for 10 days, the scale thickness sensor detects 0.5mm, and the intelligent control module starts high-pressure spray cleaning, and after 10min, the scale thickness decreases to 0.05mm, and the heat exchange efficiency recovers to 93%; Heat medium maintenance: After running for 3 months, the heat medium purity decreases to 92% and the viscosity increases to 35mPa・s, the filter and additive supplement device are started, and 2L / h of T501 antioxidant is supplemented, and after 2h, the purity recovers to 96% and the viscosity decreases to 25mPa・s. (III) Implementation effect In this embodiment, the system heat exchange efficiency is improved by 8% (from 85% to 93%), the comprehensive energy consumption is reduced by 18%, the heat conducting oil service life is extended by 30%, the manual cleaning frequency is reduced by 6 times / year, the operation and maintenance cost is reduced by 22%, and the demand for continuous and stable heat exchange of chemical production is fully met.

[0021] Example 2: Central heating and hot water heat exchange system (intermittent fluctuation condition) (I) System adaptation adjustment Heat medium: 50% ethylene glycol aqueous solution (freezing point -35℃, specific heat capacity 3.5kJ / (kg・℃)); Heat exchange unit: shell-and-tube heat exchanger (F type, heat exchange area 80 m2, pressure rating 1.6 MPa), suitable for large flow demand of heating circulating water; Actuator: variable frequency circulating pump (ISG150-200, flow rate 0-200 m3 / h, head 25 m, power 15 kW), electric heating unit replaced by steam heating coil (steam pressure 0.6 MPa, heating power 0-80 kW); Multi-parameter monitoring module: added environmental temperature sensor (DS18B20, measurement range -40-85°C, accuracy ±0.2°C), installed in an open outdoor area. (2) Fluctuating operating condition test process Test conditions: heating area 50,000 m2, heating water temperature target value 55°C, outdoor temperature fluctuation range -10°C-15°C, heating period 6:00-24:00 (intermittent load); Key test scenarios: Scenario 1: sudden outdoor temperature drop (14:00-15:00, from 5°C to -8°C, change rate 1.3°C / min); Operating condition identification: identified by the model as "fluctuating load → high load (85% rated load)", identification time 0.09 s; Load prediction: LSTM model predicts that the load will remain high for the next 10 min, with a prediction error of 3.5%; Control response: steam heating coil opening increased from 40% to 75%, circulating pump frequency increased from 30 Hz to 40 Hz (flow rate 150 m3 / h), heating water temperature stabilized at 55.1°C within 30 s, fluctuation ≤±0.3°C.

[0022] Scenario 2: low load at night (23:00-6:00, outdoor temperature -10°C, heating load 40%); Control strategy: reduce circulating pump frequency to 20 Hz (flow rate 80 m3 / h), heating power to 30 kW, heat exchange efficiency maintained at 90.5%, energy consumption reduced by 21% compared to traditional systems. (3) Long-term operation verification (3 months) Example 3: Emergency handling verification of abnormal operating conditions (1) Abnormal scenario 1: sensor failure Fault simulation: interruption of heat medium outlet temperature sensor signal; Emergency strategy: intelligent control module automatically enables redundant algorithm, calculates heat medium outlet temperature based on other parameters (heat medium inlet temperature, heat medium outlet temperature, heating power) through PID neural network (error ≤0.8°C), system continues to run without shutdown; Alarm mechanism: sound and light alarm triggered at the same time, display the location of the fault sensor, convenient for maintenance. (2) Abnormal scenario 2: sudden power failure Emergency response: the heat storage tank is equipped with a thermal insulation layer (thickness 100mm, thermal conductivity ≤0.03W / (m・K)) to maintain the temperature of the heat medium, and the one-way valve is automatically closed to prevent backflow. Restart process: after power recovery, the system automatically performs initialization detection (sensor calibration, heat medium state detection), and resumes normal operation within 5 minutes, with a heating water temperature fluctuation of ≤±1.0℃.

[0023] It should be noted that in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual such relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment. Without more limitations, the element defined by the statement "including a" does not exclude the presence of other identical elements in the process, method, article or equipment including the element.

[0024] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control, characterized in that, include: Heat medium circulation loop: used for the circulation and transportation of heat medium, including 304 stainless steel circulation pipes with a volume of 5-10m³. 3 The heat storage tank and one-way valve control the flow rate of the heat medium at 0.5-3 m / s. The heat medium is selected from heat transfer oil, ethylene glycol aqueous solution or molten salt. Heat exchange unit: installed in the heat medium circulation loop, using shell and tube type with a pressure rating of 1.6-4.0 MPa, or plate type with a heat exchange area of ​​10-100 m². 2 The heat exchanger has a heat medium channel and a heat exchange medium channel arranged in counter-current flow, and both ends are equipped with temperature and pressure detection interfaces. Multi-parameter monitoring module: integrates sensors for temperature, pressure, flow rate, viscosity, purity, water content, and scale thickness; data acquisition cycle is 0.1-1s, and transmission delay is ≤0.3s. Intelligent control module: It communicates with the multi-parameter monitoring module, has a built-in working condition identification model and a hybrid predictive control algorithm of PID neural network and LSTM, and supports offline calculation and online update; Actuators: including flow rates of 0-100m 3 A variable frequency circulating pump with a head of 10-50m, an electric three-way regulating valve, a 0-50kW adjustable power heat medium heating unit, and an air-cooled / water-cooled cooling unit. Heat medium condition optimization module: includes a 5-10μm precision filter, a vacuum dehydrator and an automatic metering pump, used for heat medium purification and performance restoration; Automatic cleaning device: Linked with the intelligent control module, it adopts 8-12MPa high-pressure spray or ultrasonic cleaning method and is equipped with a waste liquid recovery unit.

2. The high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The multi-parameter monitoring module includes a PT100 temperature sensor with a range of -50 to 300°C and an accuracy of ±0.1°C, a diffused silicon pressure sensor with a range of 0 to 2 MPa and an accuracy of ±0.5%FS, and a 0-100m... 3 / h, electromagnetic flowmeter with an accuracy of ±0.3%; 5-1000mPa・s, vibration viscosity sensor with an accuracy of ±2%; 90%-100%, near-infrared spectral purity sensor with an accuracy of ±1%; and 0-5mm, ultrasonic scale thickness sensor with an accuracy of ±0.01mm.

3. The high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The intelligent control module adopts an industrial-grade PLC controller or edge computing gateway, with a built-in Python runtime environment and a data storage unit of ≥32GB. It supports offline calculation and online algorithm updates, and the control response time is ≤1s.

4. The high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The heating unit for the heat transfer medium is either electric or steam-heated, and the cooling unit is either air-cooled or water-cooled heat exchanger. The additive replenishment device uses an automatic metering pump with a range of 0-5L / h and an accuracy of ±0.1L / h. The additives are special antioxidants, anti-scaling agents, or molten salt stabilizers for heat transfer oil.

5. The high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The intelligent control module supports Ethernet, Modbus RTU, 4G / 5G and LoRa wireless communication protocols. Data transmission is encrypted with AES-128 and has the function of resuming interrupted transmission to ensure the security and continuity of parameter transmission.

6. The high-efficiency heat exchange system for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The heat medium state optimization module uses multi-parameter linkage control: when the water content of the heat medium is ≥0.3% and the purity is ≤94%, the dehydration device and additive replenishment program are started simultaneously. The amount of additive replenishment is dynamically adjusted according to the real-time purity of the heat medium, with an adjustment accuracy of ±0.05L / h.

7. A method for efficient heat exchange through intelligent control of a heat transfer medium circulation based on the system described in any one of claims 1-6, characterized in that, Includes the following steps: S1: Multi-parameter acquisition and preprocessing: The operating parameters are acquired through the multi-parameter monitoring module, and after outlier removal by the 3σ criterion, smoothing and normalization by the moving average method, the data are transmitted to the intelligent control module. S2: Intelligent Operating Condition Recognition: Calls the trained operating condition recognition model to identify low load ≤30% of rated load, medium load 30%-70%, high load ≥70%, and fluctuating load ≥10% / min change rate, with an accuracy rate ≥95%; S3: Heat exchange load prediction: Based on the LSTM model, predict the load change trend in the next 5-10 minutes, with the model mean square error MSE ≤ 0.01; S4: Control command generation: Calculates the optimal heat medium flow rate, temperature and circulation pressure parameters through the PID neural network algorithm and converts them into control commands for the actuator; S5: Closed-loop control and status optimization: After the actuator responds to the command, the multi-parameter monitoring module provides real-time feedback data and dynamically corrects the control command; when the conditions for heat medium status optimization or automatic cleaning are triggered, the corresponding module is activated. S6: Continuous optimization: Regularly update the parameters of the operating condition identification model and prediction algorithm to improve the accuracy of control.

8. The efficient heat exchange method for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The PID neural network algorithm includes an input layer with 5 historical load data and 5 real-time parameters, a 3-layer hidden layer with 16-32 neurons per layer, and an output layer with 3 neurons. It is trained using the Adam optimizer with ≥1000 iterations and optimizes the objective function of "maximizing heat exchange efficiency and minimizing energy consumption" using gradient descent.

9. The efficient heat exchange method for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, In S5, the conditions for optimizing the state of the heat transfer medium are purity ≤92%, viscosity ≥35mPa・s or water content ≥0.5%, filter working pressure ≤0.3MPa during purification, and vacuum degree of the dehydration device ≥-0.09MPa; the conditions for automatic cleaning are scale thickness ≥0.5mm, cleaning time 5-15min, and heat transfer medium flow rate maintained at 70%-80% of the rated value during the cleaning process.

10. The efficient heat exchange method for heat transfer medium circulation based on intelligent control according to claim 1, characterized in that, The LSTM model in S3 uses a 3-layer hidden layer with 24-32 neurons per layer. It takes load data, heat medium temperature and heat exchange medium flow parameters from 5 historical time points as input, optimizes model parameters through 10-fold cross-validation, and directly outputs the prediction result when the prediction error is ≤5%. When the error is >5%, it starts the PID neural network to assist in correction.

Images