Energy cascade recovery and multi-stage heat recovery methods for high and low temperature zoned CO2 transcritical refrigeration systems

Abstract

This invention relates to the field of refrigeration and heat pump system technology, specifically disclosing a method for energy cascade recovery and multi-stage heat recovery in a high- and low-temperature zoned CO2 transcritical refrigeration system. The method includes: in the target refrigeration system, collecting the refrigerant density gradient and the temperature difference between the inlet and outlet cooling water, constructing a transient response model of the refrigerant density gradient and the temperature difference between the inlet and outlet cooling water, predicting the direction of load mutation, and adjusting the opening of the electronic expansion valve in advance to anchor the exhaust pressure to the real-time optimal value under the current load. This invention, by constructing a transient response model of the refrigerant density gradient and the temperature difference between the inlet and outlet cooling water, outputs a load mutation prediction flag in advance before the PID controller acts and generates an advance compensation amount for the electronic expansion valve, transforming the exhaust pressure regulation from a passive, lagging response to an active, feedforward pre-adjustment. This suppresses drastic fluctuations in cooling water temperature recovery, eliminates the system energy efficiency degradation caused by lag in traditional control methods, and ensures that the high-pressure side pressure is always anchored to the optimal operating point under the current load conditions.

Description

Technical Field

[0001] This invention relates to the field of refrigeration and heat pump system technology, and in particular to a method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system. Background Technology

[0002] As fluorinated refrigerants are gradually restricted, the natural working fluid CO2 has been promoted for application due to its low global warming potential. However, its transcritical cycle has a large throttling loss, and the high-temperature sensible heat at the outlet of the gas cooler has a high usable energy grade, providing a physical basis for the graded utilization of thermal energy. At the same time, different temperature levels of cooling and heating loads often exist simultaneously in scenarios such as cold chain, commercial refrigeration, and heat pump energy supply, prompting the system design to evolve from single refrigeration to combined cooling and heating.

[0003] However, existing transcritical high-pressure side gas coolers are susceptible to sudden changes in the load at the equipment's end, leading to drastic fluctuations in cooling water return temperature and frequent deviations in the optimal exhaust pressure of the heat pump. Traditional PID control suffers from severe lag, resulting in a significant decrease in the system's COP. Simultaneously, the asynchronous defrosting cycle of the evaporator causes distortion in the return gas dryness, leading to aggravated fluctuations in superheat amplitude, further weakening the effective heat exchange of the evaporator. Furthermore, the imbalance in heat distribution between the medium-temperature condenser and the high-temperature gas cooler results in the waste of residual heat in the low-temperature section (30~40℃), making it impossible for the overall energy efficiency to adaptively optimize in a closed-loop manner according to the load. To address these issues, a high-low temperature zoned CO2 transcritical refrigeration system energy cascade recovery and multi-stage heat recovery method is proposed. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, the present invention provides a method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system, which can effectively solve the problems involved in the prior art.

[0005] The objective of this invention can be achieved through the following technical solution: This invention provides a method for energy cascade recovery and multi-stage heat recovery in a high- and low-temperature zoned CO2 transcritical refrigeration system, comprising the following steps: S1. In the target refrigeration system, the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water are collected to construct a transient response model of the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water. The direction of load change is predicted, and the opening of the electronic expansion valve is adjusted in advance so that the exhaust pressure is anchored to the real-time optimal value under the current load. This effectively suppresses the impact disturbance of load change on the high-pressure side and provides a stable pressure reference for subsequent adjustment. S2. Based on the predicted direction of load change, the target value of exhaust pressure is corrected online to suppress the drastic fluctuation of cooling water temperature recovery, eliminate the COP decay caused by PID lag, and enable the high-pressure side to operate in the high-efficiency zone adapted to the load. This ensures that the high-pressure side always operates at the energy efficiency optimal point under the current load, and significantly improves the energy efficiency of the system under partial load. S3. Collect the return gas temperature and pressure of each branch, calculate the superheat variance of each branch of the evaporator, and determine whether the evaporation pressure is allowed to actively follow the branch with the highest superheat. If it exceeds the preset threshold, actively increase the pressure to compensate and make the evaporation pressure follow the branch with the highest superheat. Otherwise, maintain the original pressure, lock the superheat fluctuation within a narrow range, restore the effective heat exchange of the evaporator, eliminate the dryness distortion caused by asynchronous defrosting, and restore the heat exchange efficiency of each branch to the normal level synchronously. S4. After the high-pressure side and superheat parameters converge, establish a dynamic mapping relationship between the energy efficiency ratio and the heat recovery water temperature requirement, forming a three-level tiered allocation system to ensure that energy allocation and real-time operating conditions are accurately matched, realize the synergistic optimization of heat recovery grade and real-time system energy efficiency, and avoid mismatch between energy supply and demand. S5. Based on the three-level distribution system, a three-level heat recovery strategy is adopted to allocate the heating flow of the high-temperature section, the reheat flow of the medium-temperature section, and the preheat flow of the low-temperature section. This strategy captures the waste heat of the low-temperature section (30~40℃), eliminates the heat distribution imbalance between the condenser and the gas cooler, and realizes the orderly recovery of waste heat in the entire temperature range, so that the heat in the low-temperature section can be effectively utilized instead of being directly lost. S6. Based on the real-time load weight of each evaporator, dynamically adjust the refrigerant flow ratio of the corresponding heat recovery branch to construct a full-condition energy adaptive closed-loop strategy, so that the energy efficiency of the refrigeration system continuously locks the optimal operating range with load fluctuations, forming a global energy self-optimization closed-loop control capability, enabling the system to maintain high energy efficiency operation under changing conditions.

[0006] Preferably, S1 specifically includes: High-frequency response temperature and pressure composite sensors are deployed in the refrigerant outlet pipeline and cooling water inlet and outlet pipeline of the gas cooler in the refrigeration system. The refrigerant density gradient time-series signal and the cooling water inlet and outlet temperature difference time-series signal are collected synchronously at a sampling frequency of not less than 10Hz, providing a high-resolution data input basis for the transient response model. The refrigerant density gradient time series signal and the cooling water inlet and outlet temperature difference time series signal are input into a pre-constructed transient response model that couples the refrigerant density gradient and the cooling water inlet and outlet temperature difference. The transient response model generates a load change direction discrimination factor based on historical load change samples and outputs a prediction indicator of a sudden increase or decrease in load, enabling the system to identify load change trends in advance and transforming passive response into active prediction. Based on the prediction indicator, an advance compensation amount for the opening of the electronic expansion valve is generated before the PID controller intervenes. This advance compensation amount is based on the opening corresponding to the optimal value of the exhaust pressure under the current load, and the opening pre-adjustment action is performed in advance, which greatly shortens the response time of pressure fluctuation and suppresses overshoot.

[0007] Preferably, S1 further includes: The advance compensation is decomposed into coarse adjustment component and fine adjustment component. The coarse adjustment component determines the opening step size based on the load change amplitude, and the fine adjustment component performs continuous linear correction based on the instantaneous change rate of refrigerant density gradient, so as to realize the graded processing of disturbances of different scales and make the adjustment action and the disturbance degree accurately matched. After the pre-adjustment action is executed, the actual response curve of the exhaust pressure is continuously monitored, the dynamic deviation between the actual response curve and the target response curve is calculated, and the dynamic deviation is used as a feedback signal input to the closed-loop adjustment link of the electronic expansion valve. The internal weight coefficient of the transient response model is updated based on the dynamic deviation, so that the accuracy of load change prediction gradually improves with the accumulation of the operating cycle. The model prediction accuracy is continuously optimized through the closed-loop self-learning mechanism, so that the prediction capability is continuously enhanced with the operating time.

[0008] Preferably, S2 specifically includes: The predicted direction of load change is used as the feedforward disturbance input, which is then input into the online correction module for the exhaust pressure target value. This module has a built-in adaptive correction function based on the cooling water return temperature change rate, which enables the correction action to directly respond to the root cause of the return temperature fluctuation and improve the targeting of the feedforward control. Based on the output of the adaptive correction function, the set target value of the exhaust pressure is adjusted. The correction direction of this target value is opposite to the direction of the load change, and the correction amplitude is positively correlated with the cooling water temperature change rate, ensuring that the correction amplitude matches the disturbance level in real time and avoiding over-adjustment or under-adjustment. The corrected exhaust pressure target value is sent to the high-pressure side pressure regulating actuator. The high-pressure side pressure regulating actuator performs PID fine adjustment based on the corrected target value to maintain the high-pressure side pressure in the high-efficiency pressure range that is adapted to the load, eliminating the energy efficiency decay caused by the lag of traditional PID, and keeping the high-pressure side always locked in the high-efficiency operating range.

[0009] Preferably, S3 specifically includes: Based on the temperature and pressure sensors arranged in the return gas pipeline of each evaporator, the return gas temperature and return gas pressure values ​​of each evaporator branch are collected synchronously, and the superheat value of each branch is calculated according to the refrigerant physical property parameters, providing a real-time data basis for the evaluation of the evaporator group's operating status. Calculate the total variance of superheat values ​​for all evaporator branches and compare it with a preset superheat variance threshold to determine whether there is return gas dryness distortion caused by asynchronous defrosting cycle in the current evaporator group. Quantitatively evaluate the degree of thermal non-uniformity between branches and accurately identify the dryness distortion state. When the overall variance exceeds the preset superheat variance threshold, an evaporation pressure boosting compensation command is generated; when the overall variance does not exceed the preset superheat variance threshold, the current evaporation pressure setting remains unchanged, and the adjustment action is only initiated when intervention is necessary to avoid system oscillations caused by unnecessary disturbances.

[0010] Preferably, S3 further includes: After generating the pressure boosting compensation command, the superheat values ​​of all evaporator branches are traversed, the branch with the highest superheat is selected, and the superheat value of this branch is used as the target following reference value for the active pressure boosting of the evaporation. Compensation is carried out with the most unfavorable branch as the reference to ensure that the overall regulation takes the worst operating condition as the convergence target. Based on the difference between the target following reference value and the saturation temperature corresponding to the current evaporation pressure, the amplitude and rate of pressure increase compensation are calculated, and the electronic expansion valve or intake throttling mechanism is controlled to perform active pressure increase action, so that the evaporation pressure approaches the saturation pressure corresponding to the target following reference value, ensuring that the pressure increase process is stable and controllable, and avoiding new operational risks caused by excessive adjustment. During the pressurization process, the real-time variance of the superheat value of each branch is continuously monitored. When the real-time variance value falls below 50% of the preset superheat variance threshold, the pressurization compensation is terminated and the system enters steady-state maintenance mode, locking the evaporation pressure at the current value. This ensures that the compensation is exited in time after the variance drops to the normal range, preventing excessive pressurization from causing energy efficiency loss.

[0011] Preferably, S4 specifically includes: Once the high-pressure side exhaust pressure fluctuation rate is less than the set convergence threshold and the overall variance of evaporator superheat returns to the preset superheat variance threshold range, the energy efficiency and heat recovery dynamic mapping mechanism is activated. Three input signals, including ambient temperature, cooling water return temperature and medium-temperature condenser load rate, are collected to ensure that the system is in a stable and controllable operating state when the mapping mechanism is started, thus avoiding invalid calculations. The three input signals are input into a pre-constructed nonlinear mapping network, which outputs the relationship curve between the system's overall energy efficiency ratio and the target water supply temperature on the heat recovery side under the current operating conditions, so that the heat recovery temperature target and the system's real-time energy efficiency are accurately correlated, avoiding deviations caused by empirical settings. Based on the corresponding relationship curve, the heat recovery side is divided into a three-level tiered distribution system, including a high-temperature heat recovery zone, a medium-temperature heat recovery zone, and a low-temperature heat recovery zone. Independent flow allocation priorities and temperature target zones are set for each zone, and heat is allocated according to grade to ensure that the temperature target of each heat recovery zone matches the optimal real-time energy efficiency operating conditions.

[0012] Preferably, S5 specifically includes: Based on the high-temperature heat recovery zone, medium-temperature heat recovery zone and low-temperature heat recovery zone in the three-stage distribution system, independent three-way proportional regulating valve groups and variable frequency circulating water pumps are configured on the inlet branches of each heat recovery heat exchanger to ensure that the heat recovery flow rate of each stage can be adjusted independently and to avoid interstage coupling interference. The outlet water temperatures of the high-temperature stage heat recovery, the medium-temperature stage heat recovery, and the low-temperature stage heat recovery are collected separately. Each outlet water temperature value is compared with the corresponding temperature target range to generate opening correction signals for each level of regulating valve group, so that the outlet water temperature of each level is always controlled within its respective target range, thus maintaining the stability of heat recovery quality. Based on the opening correction signal, the flow rate of the heat-carrying fluid flowing through the high-temperature section heat exchanger, the medium-temperature section condenser, and the low-temperature section preheating heat exchanger is independently adjusted so that the outlet water temperature of each stage approaches the center value of the corresponding temperature target range, thereby realizing the on-demand distribution of heat recovery at each stage and avoiding the impact of excessive or insufficient heat recovery at a certain stage on the overall energy efficiency.

[0013] Preferably, S5 further includes: A bypass branch is set in the outlet pipe of the low-temperature preheating heat exchanger. This bypass branch guides the captured low-temperature waste heat medium to the inlet water preheating coil of the medium-temperature condenser or the anti-frost heating pipe at the bottom of the evaporator, so that the low-temperature waste heat that is originally easy to ignore can be effectively utilized and the lower limit of heat recovery can be extended. The actual heat transfer fluid temperature at the outlet of the low-temperature preheating heat exchanger is collected. If the actual heat transfer fluid temperature is higher than the inlet water temperature of the medium-temperature condenser but lower than the outlet water temperature of the medium-temperature condenser, the bypass branch regulating valve is opened to perform directional transport of low-temperature waste heat, ensuring that low-temperature waste heat is introduced only when there is heat transfer driving force, and avoiding ineffective circulation. During the directional conveying process, the heat load change rate of the medium-temperature section condenser is monitored, and the opening of the bypass branch regulating valve is adjusted synchronously according to the heat load change rate to maintain the temperature of the medium-temperature section condenser after water preheating at the preset preheating target value, so that the preheating heat supply follows the load fluctuation of the medium-temperature section in real time and ensures that the preheating temperature is always stable.

[0014] Preferably, S6 specifically includes: Collect the superheat value of each evaporator branch and the current opening value of the corresponding electronic expansion valve. Calculate the real-time load weight coefficient of each evaporator based on the degree of superheat deviation from the standard value and the size of the opening value. This allows the refrigerant distribution weight to be dynamically adjusted according to the actual load demand of each evaporator, avoiding supply and demand imbalance caused by a fixed ratio. The real-time load weight coefficient of each evaporator is used as the basis for allocation. The theoretical allocation ratio of refrigerant flow in each heat recovery branch is calculated. The flow ratio is adjusted by adjusting the speed of the refrigerant pump in each branch or the opening of the throttling element to ensure that the refrigerant supply in each branch is accurately matched with its current heat load and to eliminate oversupply or undersupply. The actual heat output of each heat recovery branch and the actual cooling capacity on the evaporator side are fed back to the transient response model and dynamic mapping mechanism to form an energy adaptive closed-loop control link with rolling optimization under all operating conditions. This enables all parts of the system to operate collaboratively under real-time feedback drive, and the overall energy efficiency remains optimal as the load changes.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. The energy cascade recovery and multi-stage heat recovery method of the high and low temperature zone CO2 transcritical refrigeration system is proposed. By constructing a transient response model of refrigerant density gradient and cooling water inlet and outlet temperature difference, the load change prediction indicator is output in advance before the PID controller is activated, and the electronic expansion valve advance compensation amount is generated. This changes the exhaust pressure regulation from passive lag response to active feedforward pre-adjustment, suppresses the drastic fluctuation of cooling water temperature recovery, eliminates the system energy efficiency decay caused by lag in traditional control methods, and keeps the high pressure side pressure always anchored at the optimal operating point under the current load conditions.

[0016] 2. The energy cascade recovery and multi-stage heat recovery method of the high and low temperature zoned CO2 transcritical refrigeration system collects the superheat of each branch of the evaporator and calculates the overall variance. The variance threshold is used as the quantitative judgment basis for the distortion of return gas dryness. When the superheat distribution is uneven due to the asynchronous defrosting cycle, the evaporation pressure is actively made to approach the saturation pressure corresponding to the branch with the highest superheat, locking the superheat fluctuation within a narrow range, restoring the effective heat exchange of the evaporator, and eliminating the adverse effects of asynchronous defrosting on the heat exchange performance of the evaporator.

[0017] 3. The energy cascade recovery and multi-stage heat recovery method of the high and low temperature zoned CO2 transcritical refrigeration system establishes a dynamic mapping relationship between the system energy efficiency ratio and the heat recovery water temperature requirement after the dual parameters on the high pressure side and the evaporator side converge. The heat recovery side is divided into three-stage cascade distribution intervals of high temperature, medium temperature and low temperature, and independent temperature target range and flow distribution priority are set for each level, so that the energy allocation corresponds accurately with the real-time operating conditions. Attached Figure Description

[0018] Figure 1This is a schematic diagram of the working process of the energy cascade recovery and multi-stage heat recovery method of the high and low temperature zoned CO2 transcritical refrigeration system of the present invention; Figure 2 This is a schematic diagram of the method flow for the energy cascade recovery and multi-stage heat recovery method of the high and low temperature zoned CO2 transcritical refrigeration system of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] Example 1, please refer to Figure 1 , Figure 2 This invention provides a technical solution: a method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system, comprising the following steps: In the target refrigeration system, the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water are collected to construct a transient response model of the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water. This model predicts the direction of load change and adjusts the opening of the electronic expansion valve in advance so that the exhaust pressure is anchored to the real-time optimal value under the current load. This effectively suppresses the impact disturbance of load change on the high-pressure side and provides a stable pressure reference for subsequent adjustment. In some embodiments, a temperature-pressure composite sensor with a response time of no more than 0.5 seconds is installed in the refrigerant outlet pipe and the cooling water inlet and outlet pipes of the gas cooler. The temperature measurement accuracy of the sensor is ±0.15℃ (including calibration error), and the pressure measurement accuracy is ±0.25%FS (full scale of 15MPa). The sensor synchronously collects four signals: refrigerant outlet temperature, outlet pressure, cooling water inlet temperature, and outlet temperature at a fixed sampling frequency of 12Hz. After calculation by the embedded processor's internal property library, a refrigerant density gradient time-series signal and a cooling water inlet and outlet temperature difference time-series signal are generated. The sensor signal line uses a twisted-pair shielded cable and is connected to a 24-bit analog-to-digital converter channel. The sampled data is stored in a circulating buffer after median filtering (window length of 5). The depth of the circulating buffer is set to 300 sampling points, corresponding to a 25-second historical data window. All sensors must be calibrated on-site at both zero point and full scale after the first power-on or after a major overhaul. The calibration cycle shall not exceed 90 calendar days. In some embodiments, the two collected time-series signals are input into a pre-constructed transient response model. This transient response model adopts a nonlinear autoregressive network structure with external input, and has 18 hidden layer nodes. The training sample set is derived from 120 typical load change condition segments extracted from the actual operation records of the system, covering a load change rate ranging from +8% to -12% per minute. The model outputs a load change direction discrimination factor. When the discrimination factor is greater than +0.6 for three consecutive sampling periods, it is judged as a load surge; when it is less than -0.6 for three consecutive sampling periods, it is judged as a load drop; otherwise, it is judged as a steady-state or slowly changing condition, and no prediction flag is triggered. The model operation cycle is consistent with the sampling cycle, that is, a forward calculation is completed every 83.3ms. After the prediction flag is generated, the electronic expansion valve advance compensation amount is immediately calculated: based on the base opening degree corresponding to the exhaust pressure approaching the optimal value under the current load, the coarse adjustment component is determined according to the ratio of 0.8% opening step size for every 1% of the load change amplitude, and the fine adjustment component is based on the instantaneous change rate of the refrigerant density gradient every 0.05kg / m³. 3 / s corresponds to a 0.3% opening degree for linear correction. The calculation of the advance compensation amount and the issuance of the command are completed within 40ms after the prediction flag is generated, ensuring that the advance adjustment action intervenes before the integral saturation response of the PID controller. In some embodiments, after the advance compensation command is issued to the stepper motor driver of the electronic expansion valve, the valve body actuator completes the entire opening pre-adjustment stroke within 150ms. After execution, the actual exhaust pressure response curve is continuously monitored at a sampling rate of 2Hz. The actual pressure value at each monitoring point is compared with the corresponding interpolation point on the target response curve to calculate the dynamic deviation. The target response curve is generated in real time according to the current load change rate, with a time constant of 2.5 seconds. The overshoot is limited to no more than 3% of the target value. When the absolute value of the dynamic deviation has been continuously monitored for six cycles... When the pressure exceeds 0.08 MPa, the pre-adjustment accuracy is deemed insufficient. This deviation is then used as a negative feedback signal input to the additional closed-loop adjustment stage of the electronic expansion valve. The proportional gain of this stage is 0.6, and the integral time is 2.0 seconds. This stage performs a secondary fine-tuning of the valve opening. Simultaneously, after normalization, the dynamic deviation is used to update the output layer connection weights of the transient response model with a learning rate of 0.02. Specifically, the update cycle is once every 50 effective prediction actions. After accumulating no less than 200 effective prediction actions, the model's accuracy in judging the direction of load change gradually improves.

[0021] Based on the predicted direction of load change, the target value of exhaust pressure is corrected online to suppress drastic fluctuations in cooling water temperature recovery, eliminate COP decay caused by PID lag, and enable the high-pressure side to operate in the high-efficiency zone adapted to the load. This ensures that the high-pressure side always operates at the energy efficiency optimal point under the current load, and significantly improves the energy efficiency of the system under partial load. In some embodiments, the predicted direction of load surge is used as the feedforward disturbance input to the online correction module for the exhaust pressure target value. This module embeds an adaptive correction function based on the cooling water return temperature change rate. After receiving the load surge or drop flag from the transient response model output, the correction module simultaneously reads the real-time return temperature change rate value collected by the cooling water return temperature sensor at a sampling rate of 2Hz. When a load surge is determined, the adaptive correction function outputs a negative correction amount, proportionally reducing the exhaust pressure target value. The reduction magnitude is based on the return temperature change rate at 0.5℃ / min. The ratio should be determined by 0.15MPa; when a sudden load drop is detected, a positive correction amount is output, and the upward adjustment value is determined according to the same ratio. The target value of the corrected exhaust pressure is limited to the safe operating range of the high-pressure side pressure regulating actuator, that is, not lower than 7.2MPa and not higher than 10.5MPa, to avoid excessive correction amount leading to compressor exhaust temperature exceeding the limit or oil film failure. The correction command is calculated and sent to the high-pressure side pressure regulating actuator within 80ms after the prediction mark is generated. The actuator includes a cooperative control unit for the variable frequency compressor and the high-pressure electronic expansion valve. In some embodiments, the high-pressure side pressure regulating actuator performs PID fine-tuning based on the corrected exhaust pressure target value. The proportional gain of the PID controller is set to 1.2, the integral time to 3.5 seconds, the derivative time to 0.8 seconds, and the control cycle to 200ms, matching the response time of the variable frequency compressor speed regulation loop. Within the first control cycle after receiving the corrected target value, the actuator uses the deviation between the current measured exhaust pressure and the corrected target value as the adjustment input, and outputs the frequency adjustment amount of the compressor inverter and the additional adjustment amount of the high-pressure electronic expansion valve. The opening correction is increased, with the frequency adjustment step limited to no more than 2Hz each time, and the high-pressure electronic expansion valve opening correction step limited to no more than 1.5% of the full stroke each time, to suppress pressure oscillations caused by over-adjustment of the actuator. When the measured exhaust pressure value enters the deviation band of the correction target value ±0.12MPa and is maintained for 5 control cycles (i.e. 1 second), the PID regulation switches to steady-state maintenance mode, limiting the integral term to within ±5% of the output range to prevent small deviations from accumulating and inducing periodic fluctuations, and ensuring that the high-pressure side pressure is continuously anchored within the high-efficiency pressure range that is adapted to the load.

[0022] Collect the return gas temperature and pressure of each branch, calculate the superheat variance of each branch of the evaporator to determine whether the evaporation pressure is allowed to actively follow the branch with the highest superheat. If it exceeds the preset threshold, actively increase the pressure to compensate and make the evaporation pressure follow the branch with the highest superheat. Otherwise, maintain the original pressure, lock the superheat fluctuation within a narrow range, restore the effective heat exchange of the evaporator, eliminate the dryness distortion caused by asynchronous defrosting, and make the heat exchange efficiency of each branch synchronously return to the normal level. In some embodiments, temperature and pressure sensors arranged in the return gas pipelines of each evaporator are used to synchronously collect the return gas temperature and pressure values ​​of each evaporator branch. Specifically, platinum resistance temperature sensors and piezoresistive pressure sensors with a response time of no more than 0.5 seconds are installed at the straight sections of the return gas pipelines of each evaporator branch. After the sensor signals are converted from analog to digital by 24-bit, the embedded processor calls the CO2 property library to calculate the superheat value corresponding to each branch. The superheat calculation cycle is consistent with the system control cycle, which is 200ms. After the superheat values ​​of all evaporator branches are collected, their overall variance is calculated to characterize the dispersion of superheat between branches. The preset superheat variance threshold is 2.0K. 2 This threshold, derived from extensive experimental calibration, reflects the maximum allowable superheat distribution dispersion within a normal defrosting cycle. The real-time calculated overall variance is compared to this threshold; when the variance exceeds 2.0K... 2 If the system determines that the current evaporator group has a return gas dryness distortion caused by asynchronous defrosting cycles, meaning that some branches have liquid carryover in the return gas or excessively high superheat due to inconsistent defrosting progress, an evaporator pressure boosting compensation command is immediately generated; if the variance value does not exceed 2.0K... 2 If the evaporator unit is found to be operating normally, the current evaporation pressure setpoint will be maintained and the pressure boosting action will not be triggered. In some embodiments, after generating the pressure boosting compensation command, the superheat values ​​of all evaporator branches are iterated, and the branch with the highest superheat is selected. The superheat value of this branch is then used as the target reference value for the active pressure boosting of the evaporator. For example, if there are four evaporator branches with measured superheat values ​​of 1.2K, 3.8K, 5.5K, and 2.3K, the branch corresponding to 5.5K is selected as the target reference branch. Based on the difference between the saturation temperature corresponding to this target reference value and the saturation temperature at the current evaporation pressure, the amplitude and rate of the pressure boosting compensation are calculated. The pressure boosting rate is limited to no more than [a certain value]. 0.08MPa / s, to prevent the risk of liquid carry-in in the compressor suction due to excessive pressure rise. The total pressure rise amplitude does not exceed 12% of the current evaporation pressure setting. The electronic expansion valve or suction throttling mechanism is controlled to perform active pressure rise action, so that the evaporation pressure gradually approaches the saturation pressure corresponding to the target following reference value. During the pressure rise process, the deviation between the current evaporation pressure and the target saturation pressure is recalculated every control cycle (200ms), and the opening of the throttling mechanism is adjusted proportionally. The opening correction step is limited to no more than 1.2% of the full stroke each time to ensure that the pressure rise process is stable and controllable and to avoid pressure oscillations that cause system protection shutdown. In some embodiments, during the boost operation, the real-time variance of the superheat value of each branch is continuously monitored at a period of 200ms. When the real-time variance value falls below a preset superheat variance threshold of 2.0K, the system will take action. 2 50%, that is, less than 1.0K 2When the pressure rise compensation is terminated, the system switches to steady-state maintenance mode. In steady-state maintenance mode, the evaporating pressure is locked at the current value, and the integral term of the electronic expansion valve or suction throttling mechanism is limited to within ±3% of the output range to prevent small deviations from accumulating and causing slow pressure drift. If, after locking, the overall variance of superheat remains at 1.0K for 30 consecutive control cycles (i.e., 6 seconds), the system will maintain this value. 2 If the superheat of each branch is within the allowable operating range of 3.0K to 7.0K, then the pressure compensation is confirmed to be successful, and the current evaporation pressure setpoint will be maintained for continued operation; if the variance value rises back to 1.0K after locking, the pressure compensation will be considered successful. 2 Above but not exceeding 2.0K 2 If this occurs, the fine-tuning action in maintenance mode is triggered, making minor corrections to the evaporation pressure at a rate not exceeding 0.02 MPa / s, with the correction magnitude limited to no more than 3% of the current set value, until the variance value falls back to 1.0 K. 2 The following method aims to achieve narrow-band locking of the superheat distribution in the evaporator group, restore the effective heat exchange of the evaporator, and eliminate the heat exchange attenuation caused by asynchronous defrosting.

[0023] Example 2, as Figure 1 , Figure 2 As shown, based on Example 1, the present invention provides a technical solution: after the high-pressure side and superheat degree converge, a dynamic mapping relationship between the energy efficiency ratio and the heat recovery water temperature requirement is established to form a three-level tiered allocation system, ensuring that energy allocation and real-time operating conditions are accurately matched, realizing the synergistic optimization of heat recovery grade and real-time system energy efficiency, and avoiding mismatch between energy supply and demand. In some embodiments, the high-pressure side exhaust pressure fluctuation rate is maintained within ±0.8% for 30 consecutive control cycles (i.e., 6 seconds), and the overall variance of evaporator superheat remains below 1.0K. 2Subsequently, the dynamic mapping mechanism between energy efficiency and heat recovery is activated. The embedded processor simultaneously acquires three input signals: an ambient temperature sensor deployed outside the unit reads the outdoor ambient temperature at a sampling rate of 0.5Hz; a platinum resistance temperature sensor on the cooling water return main reads the return water temperature at a sampling rate of 2Hz; and the real-time load rate is obtained by integrating the enthalpy difference between the refrigerant side inlet and outlet of the intermediate-temperature condenser with the flow rate on the heat transfer side. After normalization, the three signals are input into a pre-constructed nonlinear mapping network. This network adopts a three-layer feedforward structure with 12 hidden layer nodes. The output layer generates the corresponding curve of the system's comprehensive energy efficiency ratio and the target supply water temperature on the heat recovery side under the current operating conditions. The forward calculation of the mapping network... The cycle is set to 2 seconds to match the response inertia of the regulating valve on the heat recovery side. Based on the corresponding relationship curve, the heat recovery side is divided into three tiered distribution ranges: high temperature, medium temperature, and low temperature. The target temperature range for the high temperature stage is set to 75 to 85 degrees Celsius, corresponding to the high-temperature heating demand of the gas cooler outlet section. The medium temperature stage is set to 45 to 55 degrees Celsius, corresponding to the reheating demand of the medium-temperature condenser. The low temperature stage is set to 30 to 40 degrees Celsius, corresponding to the preheating demand of the final stage of the high-pressure section. At the same time, independent flow allocation priorities are set for each stage: the priority coefficient for the high temperature stage is 0.35, for the medium temperature stage it is 0.40, and for the low temperature stage it is 0.25, ensuring that energy allocation accurately corresponds to real-time operating conditions. In some embodiments, after the three-stage distribution system is established, the independently configured electric three-way proportional regulating valves and variable frequency circulating water pumps on the inlet branches of each stage of the heat recovery heat exchanger are immediately put into closed-loop control. The response time of each of the three regulating valves is no more than 2 seconds, and the control cycle is consistent with the forward calculation cycle of the mapping network, which is 2 seconds. The embedded processor reads the outlet water temperature of the high-temperature stage heat recovery, the outlet water temperature of the medium-temperature stage condenser, and the outlet water temperature of the low-temperature stage preheating heat exchanger, respectively, and compares each measured value with the center value of the corresponding temperature target interval, generating an independent opening correction signal for each regulating valve. The correction signal is processed by the PID controller and output to the valve body actuator. The proportional gain is uniformly set to 1.5, the integral time is set to 4.0 seconds, and the derivative time is set to 1.2 seconds. The speed of each variable frequency circulating water pump is simultaneously adjusted in coordination according to the opening of the corresponding regulating valve. The speed adjustment step is limited to no more than 2% of the rated speed each time to prevent flow. To mitigate the oscillations in heat recovery outlet water temperature caused by sudden changes in heat volume, a bypass branch is installed at the outlet pipe of the low-temperature preheating heat exchanger. This bypass branch, via an electric regulating valve, selectively delivers the captured 30-40°C waste heat medium to the inlet preheating coil of the medium-temperature condenser or the anti-frost heating pipe at the bottom of the evaporator. The opening criterion for the bypass branch's electric regulating valve is that the measured outlet temperature of the low-temperature preheating heat exchanger is at least 1.5°C higher than the inlet water temperature of the medium-temperature condenser and at least 2.0°C lower than the outlet water temperature of the medium-temperature condenser. When this condition is met, the valve gradually opens at a rate not exceeding 1.8% of the full stroke per control cycle. Simultaneously, the heat load change rate of the medium-temperature condenser is monitored in real time, and the bypass opening is dynamically adjusted based on this change rate to stabilize the preheated inlet water temperature of the medium-temperature condenser within a preset preheating target value ±0.8°C. The flow distribution of each branch is recalculated every 5 seconds to ensure that the outlet water temperature of each stage continuously approaches the center value of the corresponding temperature target range.

[0024] Based on the three-level distribution system, a three-level heat recovery strategy is adopted to allocate the heating flow of the high-temperature section, the reheat flow of the medium-temperature section, and the preheat flow of the low-temperature section. This strategy captures the waste heat of the low-temperature section (30~40℃), eliminates the heat distribution imbalance between the condenser and the gas cooler, realizes the orderly recovery of waste heat in the entire temperature range, and makes the heat of the low-temperature section effectively utilized rather than directly lost. In some embodiments, based on the high-temperature stage heat recovery zone, the medium-temperature stage heat recovery zone, and the low-temperature stage heat recovery zone in the three-stage cascade distribution system, independent three-way proportional regulating valve groups and variable frequency circulating water pumps are configured on the inlet branches of each stage heat recovery heat exchanger. The response time of each regulating valve is no more than 2 seconds, and the control cycle is set to 2 seconds, consistent with the forward calculation cycle of the energy efficiency mapping network. The embedded processor reads the outlet water temperature of the high-temperature stage heat recovery, the outlet water temperature of the medium-temperature stage condenser, and the outlet water temperature of the low-temperature stage preheating heat exchanger, respectively. Each temperature sensor is a platinum resistance type with a measurement accuracy of ±0.15 degrees Celsius. The processor compares each measured outlet water temperature value with the center value of the corresponding temperature target zone, where the center value of the high-temperature stage is 80 degrees Celsius, the center value of the medium-temperature stage is 50 degrees Celsius, and the center value of the low-temperature stage is 35 degrees Celsius. After calculating the temperature deviation, an independent opening correction signal for each regulating valve is generated. The correction signal is then processed by P After the ID controller calculates, the output is sent to the valve body actuator. The PID parameters are uniformly set as proportional gain 1.5, integral time 4.0 seconds, and derivative time 1.2 seconds. The speed of each variable frequency circulating water pump is simultaneously adjusted in coordination according to the opening of the corresponding regulating valve. When the regulating valve opening increases, the pump speed increases accordingly to increase the flow of heat transfer fluid. When the regulating valve opening decreases, the pump speed decreases accordingly. The speed adjustment step size is limited to no more than 2% of the rated speed each time to prevent the heat recovery outlet water temperature from oscillating due to sudden changes in flow. The flow distribution of each branch is recalculated every 5 seconds to keep the outlet water temperature of each stage close to the center value of the corresponding temperature target range, adapting to the slow changes in the external environment and load. Through independent closed-loop control, the heat recovery outlet water temperature of each stage is stably maintained within its respective target range: the high temperature stage is stable in the range of 75 to 85 degrees Celsius, the medium temperature stage is stable in the range of 45 to 55 degrees Celsius, and the low temperature stage is stable in the range of 30 to 40 degrees Celsius. In some embodiments, a bypass branch is provided in the outlet pipe of the low-temperature preheating heat exchanger. This bypass branch, via an electric regulating valve, selectively delivers the captured low-temperature waste heat medium to the inlet water preheating coil of the medium-temperature condenser or the anti-frost heating pipe at the bottom of the evaporator. The embedded processor collects the actual heat transfer fluid temperature at the outlet of the low-temperature preheating heat exchanger at a sampling rate of 2 Hz, and simultaneously collects the inlet and outlet water temperatures of the medium-temperature condenser. The opening of the bypass branch electric regulating valve requires the simultaneous fulfillment of three criteria: first, the measured outlet temperature of the low-temperature preheating heat exchanger is at least 1.5 degrees Celsius higher than the inlet water temperature of the medium-temperature condenser; second, the measured outlet temperature is at least 2.0 degrees Celsius lower than the outlet water temperature of the medium-temperature condenser; and third, the real-time load rate of the medium-temperature condenser is higher than a preset demand threshold. When all three conditions are met simultaneously, the valve... The bypass valve opens gradually at a rate not exceeding 1.8% of the full stroke per control cycle to prevent the water temperature in the medium-temperature section from being too fast. After the bypass valve opens, the heat load change rate of the medium-temperature section condenser is monitored in real time, and the bypass opening is dynamically adjusted according to the heat load change rate: when the heat load increases, the bypass opening is increased to supplement more preheating heat, and when the heat load decreases, the bypass opening is decreased to reduce the supply of preheating heat. The control objective is to keep the temperature of the medium-temperature section condenser after preheating stable within the preset preheating target value ±0.8 degrees Celsius. When the temperature difference driving force between the medium-temperature section condenser inlet water temperature and the low-temperature section outlet temperature is insufficient, or when the heat load rate of the medium-temperature section drops below the demand threshold, the bypass regulating valve closes gradually at the same rate to stop the directional delivery of low-temperature waste heat and ensure that low-temperature waste heat is captured and utilized only when there are effective utilization conditions. In some embodiments, a platinum resistance temperature sensor and a piezoresistive pressure sensor with a response time of no more than 0.5 seconds are installed at the straight section of the return gas pipeline of each evaporator branch. After the sensor signals are converted from analog to digital by 24 bits, the embedded processor calls the carbon dioxide property library at a period of 200 milliseconds to calculate the superheat value corresponding to each branch. After the superheat values ​​of all evaporator branches have been collected, the processor calculates its overall variance and compares this variance value with a preset superheat variance threshold of 2.0K. 2 For comparison, when the variance value exceeds 2.0K... 2When the current evaporator group is found to have a return gas dryness distortion caused by asynchronous defrosting cycle, an evaporator pressure boosting compensation command is immediately generated. After the boosting compensation command is generated, the processor traverses the superheat values ​​of all evaporator branches, selects the branch with the highest superheat as the target following reference branch, and calculates the boosting compensation amplitude and rate based on the difference between the saturation temperature corresponding to the superheat value of the target following reference branch and the saturation temperature under the current evaporation pressure. The boosting rate is limited to no more than 0.08 MPa per second to prevent the risk of liquid carryover in the compressor suction due to excessively rapid boosting; the total boosting amplitude does not exceed 12% of the current evaporation pressure setpoint, and the electronic expansion valve is controlled to perform active boosting action, so that the evaporation pressure gradually approaches the saturation pressure corresponding to the target following reference value. In each control cycle, the deviation between the current evaporation pressure and the target saturation pressure is recalculated, and the opening of the throttling mechanism is adjusted proportionally. The opening correction step is limited to no more than 1.2% of the full stroke each time to ensure that the boosting process is stable and controllable. When the real-time variance value falls below 1.0K, the boosting compensation is completed. 2 The pressure boost compensation is terminated and the system switches to steady-state maintenance mode. The evaporation pressure is locked at the current value, and the integral term of the electronic expansion valve is limited to ±3% of the output range to prevent small deviations from accumulating and causing slow pressure drift. After locking, if the overall variance of superheat remains at 1.0K for 30 consecutive control cycles (6 seconds), the system will maintain this level. 2 If the superheat of each branch is within the allowable operating range of 3.0K to 7.0K, then the pressure compensation is confirmed to be successful, and the current evaporation pressure setting will be maintained for continued operation.

[0025] Based on the real-time load weight of each evaporator, the refrigerant flow ratio of the corresponding heat recovery branch is dynamically adjusted to construct an energy adaptive closed-loop strategy for all operating conditions. This enables the energy efficiency of the refrigeration system to continuously lock the optimal operating range as the load fluctuates, forming a global energy self-optimization closed-loop control capability, so that the system can continuously maintain high energy efficiency under changing operating conditions. In some embodiments, the superheat value of each evaporator branch and the current opening value of the corresponding electronic expansion valve are collected. Based on the absolute deviation of the superheat from the standard value of 3.0K to 7.0K (the center value of the allowable working range of 5.0K) and the percentage of the current opening value of the electronic expansion valve to the full stroke, the two are multiplied and divided by the sum of the products of each branch to calculate the real-time load weight coefficient of each evaporator branch. The embedded processor uses the real-time load weight coefficient of each evaporator branch as the basis for refrigerant flow distribution, calculates the theoretical distribution ratio of refrigerant flow for each heat recovery branch, and performs flow ratio adjustment by adjusting the speed of the refrigerant pump or the opening of the throttling element of each branch. The flow ratio adjustment cycle is consistent with the recalculation cycle of the flow distribution on the heat recovery side, that is, it is executed once every 5 seconds. The speed adjustment step is limited to no more than 2% of the rated speed each time, and the throttling element opening correction step is limited to no more than 1.2% of the full stroke each time to ensure that the refrigerant supply is accurately matched with the real-time load of each evaporator. In some embodiments, the actual heat output of each heat recovery branch and the actual cooling capacity on the evaporator side are fed back to the transient response model and the dynamic mapping mechanism. The actual heat output of each heat recovery branch is obtained by multiplying the mass flow rate of the heat transfer fluid, its specific heat capacity, and the temperature difference between the inlet and outlet water. The actual cooling capacity on the evaporator side is obtained by multiplying the mass flow rate of the refrigerant and the enthalpy difference between the inlet and outlet of the evaporator. Both types of heat feedback values ​​are updated every 5 seconds and synchronously input to a nonlinear autoregressive network model with external input and a three-layer feedforward nonlinear mapping network to update the internal weight coefficients of the transient response model and the network parameters of the dynamic mapping mechanism. The learning rate of the weight update of the transient response model is set to 0.02, and an update is performed every 50 effective prediction actions. The parameter update cycle of the dynamic mapping network is consistent with the control cycle on the heat recovery side, that is, an online correction is performed every 2 seconds. This constitutes a closed-loop control link for full-condition rolling optimization from load prediction, pressure correction, superheat compensation, energy efficiency mapping, heat recovery distribution to heat feedback, so that the refrigeration system continuously locks the optimal operating range of comprehensive energy efficiency when the load changes and the heat demand changes.

[0026] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for energy cascade recovery and multi-stage heat recovery in a high- and low-temperature zoned CO2 transcritical refrigeration system, characterized in that, Includes the following steps: S1. In the target refrigeration system, collect the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water, construct a transient response model of the refrigerant density gradient and the temperature difference between the inlet and outlet of the cooling water, predict the direction of load change, and adjust the opening of the electronic expansion valve in advance so that the exhaust pressure is anchored to the real-time optimal value under the current load. S2. Based on the predicted direction of load change, the target value of exhaust pressure is corrected online to suppress the drastic fluctuation of cooling water temperature recovery, eliminate the COP decay caused by PID lag, and enable the high-pressure side to operate in the high-efficiency zone that is adapted to the load. S3. Collect the return gas temperature and pressure of each branch, calculate the superheat variance of each branch of the evaporator, and determine whether the evaporation pressure is allowed to actively follow the branch with the highest superheat. If it exceeds the preset threshold, actively increase the pressure to compensate and make the evaporation pressure follow the branch with the highest superheat. Otherwise, maintain the original pressure and lock the superheat fluctuation within a narrow range to restore the effective heat exchange of the evaporator. S4. After the high-pressure side and superheat parameters converge, establish a dynamic mapping relationship between the energy efficiency ratio and the heat recovery water temperature requirement to form a three-level cascade distribution system. S5. Based on the three-level distribution system, a three-level heat recovery strategy is adopted to allocate the heating flow of the high-temperature section, the reheat flow of the medium-temperature section, and the preheat flow of the low-temperature section to capture the waste heat of the low-temperature section. S6. Based on the real-time load weight of each evaporator, dynamically adjust the refrigerant flow ratio of the corresponding heat recovery branch to construct a full-condition energy adaptive closed-loop strategy, so that the energy efficiency of the refrigeration system continuously locks the optimal operating range as the load fluctuates.

2. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S1 specifically includes: High-frequency response temperature and pressure composite sensors are deployed in the refrigerant outlet pipeline of the gas cooler and the inlet and outlet pipelines of the cooling water in the refrigeration system to synchronously collect the refrigerant density gradient time-series signal and the cooling water inlet and outlet temperature difference time-series signal at a sampling frequency of not less than 10Hz. The refrigerant density gradient time series signal and the cooling water inlet and outlet temperature difference time series signal are input into a pre-constructed transient response model coupled with the refrigerant density gradient and the cooling water inlet and outlet temperature difference. The transient response model is trained based on historical load change samples to generate a load change direction discrimination factor and outputs a prediction indicator of a sudden increase or decrease in load. Based on the prediction indicator, an advance compensation amount for the opening of the electronic expansion valve is generated before the PID controller intervenes. This advance compensation amount is based on the opening corresponding to the optimal value of the exhaust pressure under the current load, and the opening pre-adjustment action is performed in advance.

3. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 2, characterized in that: S1 further includes: The advance compensation amount is decomposed into coarse adjustment component and fine adjustment component. The coarse adjustment component determines the opening step size based on the load change amplitude, and the fine adjustment component performs continuous linear correction based on the instantaneous change rate of refrigerant density gradient. After the pre-adjustment action is executed, the actual response curve of the exhaust pressure is continuously monitored, the dynamic deviation between the actual response curve and the target response curve is calculated, and the dynamic deviation is used as a feedback signal to input the closed-loop regulation link of the electronic expansion valve. The internal weight coefficients of the transient response model are updated based on the dynamic deviation.

4. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S2 specifically includes: The predicted direction of load change is used as the feedforward disturbance input, and the target value of exhaust pressure is input to the online correction module. This module has a built-in adaptive correction function based on the cooling water return temperature change rate. Based on the output of the adaptive correction function, the set target value of the exhaust pressure is adjusted. The correction direction of this target value is opposite to the direction of the load change, and the correction amplitude is positively correlated with the cooling water temperature change rate. The corrected exhaust pressure target value is sent to the high-pressure side pressure regulating actuator, which performs PID fine regulation based on the corrected target value to maintain the high-pressure side pressure within the high-efficiency pressure range adapted to the load.

5. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S3 specifically includes: Based on the temperature and pressure sensors arranged in the return gas pipeline of each evaporator, the return gas temperature and return gas pressure values ​​of each evaporator branch are collected synchronously, and the superheat value corresponding to each branch is calculated according to the refrigerant physical property parameters. Calculate the total variance of the superheat values ​​of all evaporator branches, and compare the total variance value with the preset superheat variance threshold to determine whether there is a return gas dryness distortion caused by the asynchronous defrosting cycle in the current evaporator group. When the overall variance exceeds the preset superheat variance threshold, an evaporation pressure boosting compensation command is generated; when the overall variance does not exceed the preset superheat variance threshold, the current evaporation pressure setting remains unchanged.

6. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 5, characterized in that: S3 further includes: After generating the pressure boosting compensation command, the superheat values ​​of all evaporator branches are traversed, the branch with the highest superheat is selected, and the superheat value of that branch is used as the target following reference value for the active pressure boosting of the evaporation. Based on the difference between the target following reference value and the saturation temperature corresponding to the current evaporation pressure, the amplitude and rate of pressure boost compensation are calculated, and the electronic expansion valve or intake throttling mechanism is controlled to perform active pressure boosting action, so that the evaporation pressure approaches the saturation pressure corresponding to the target following reference value. During the pressurization process, the real-time variance of the superheat value of each branch is continuously monitored. When the real-time variance value falls below 50% of the preset superheat variance threshold, the pressurization compensation is terminated and the system enters steady-state maintenance mode, locking the evaporation pressure at the current value.

7. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S4 specifically includes: Once the high-pressure side exhaust pressure fluctuation rate is less than the set convergence threshold and the overall variance of evaporator superheat returns to the preset superheat variance threshold range, the energy efficiency and heat recovery dynamic mapping mechanism is activated, and three input signals including ambient temperature, cooling water return temperature and medium-temperature condenser load rate are collected. The three input signals are input into a pre-constructed nonlinear mapping network, and the corresponding curve between the system's comprehensive energy efficiency ratio and the target water supply temperature on the heat recovery side under the current operating conditions is output. Based on the corresponding relationship curve, the heat recovery side is divided into a three-level tiered allocation system, including a high-temperature heat recovery zone, a medium-temperature heat recovery zone, and a low-temperature heat recovery zone, and independent flow allocation priorities and temperature target zones are set for each zone.

8. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S5 specifically includes: Based on the high-temperature heat recovery zone, medium-temperature heat recovery zone and low-temperature heat recovery zone in the three-stage distribution system, independent three-way proportional regulating valve groups and variable frequency circulating water pumps are configured on the inlet branches of each heat recovery heat exchanger. The outlet water temperatures of the high-temperature stage heat recovery, the medium-temperature stage heat recovery, and the low-temperature stage heat recovery are collected separately, and each outlet water temperature value is compared with the corresponding temperature target range to generate the opening correction signal of each level of regulating valve group. Based on the opening correction signal, the flow rate of the heat transfer fluid flowing through the high-temperature section heat exchanger, the medium-temperature section condenser, and the low-temperature section preheating heat exchanger is independently adjusted so that the outlet water temperature of each stage approaches the center value of the corresponding temperature target range.

9. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 8, characterized in that: The S5 also includes: A bypass branch is installed in the outlet pipe of the low-temperature preheating heat exchanger. This bypass branch guides the captured low-temperature waste heat medium to the inlet water preheating coil of the medium-temperature condenser or the anti-frost heating pipe at the bottom of the evaporator. Collect the actual heat transfer fluid temperature at the outlet of the low-temperature preheating heat exchanger. If the actual heat transfer fluid temperature is higher than the inlet water temperature of the medium-temperature condenser but lower than the outlet water temperature of the medium-temperature condenser, open the bypass branch regulating valve to perform directional transport of low-temperature waste heat. During the directional conveying process, the heat load change rate of the medium-temperature section condenser is monitored, and the opening of the bypass branch regulating valve is adjusted synchronously according to the heat load change rate to maintain the temperature of the medium-temperature section condenser after water preheating at the preset preheating target value.

10. The method for energy cascade recovery and multi-stage heat recovery in a high-low temperature zoned CO2 transcritical refrigeration system according to claim 1, characterized in that: S6 specifically includes: Collect the superheat value of each evaporator branch and the current opening value of the corresponding electronic expansion valve, and calculate the real-time load weighting coefficient of each evaporator according to the degree of superheat deviation from the standard value and the size of the opening value. The real-time load weight coefficient of each evaporator is used as the basis for allocation. The theoretical allocation ratio of refrigerant flow in each heat recovery branch is calculated, and the flow ratio is adjusted by adjusting the speed of the refrigerant pump in each branch or the opening of the throttling element. The actual heat output of each heat recovery branch and the actual cooling capacity on the evaporator side are fed back to the transient response model and dynamic mapping mechanism to form an energy adaptive closed-loop control link with full-condition rolling optimization.

Images