An energy consumption optimization process control method and system for an industrial cold and hot temperature control unit

Abstract

The application discloses an energy consumption optimization process control method and system of an industrial cold and hot temperature control unit, comprising: obtaining process control data and a process temperature program of an industrial temperature control process; determining a dynamic temperature tolerance domain corresponding to a current process stage; determining a signed residual heat state representing unreleased cold or heat; predicting future process object temperatures and process energy consumption corresponding to a candidate process control sequence; performing rolling optimization with the goal of reducing process energy consumption and cold and hot switching cost to obtain a target process control sequence; determining a residual heat release process and a starting condition of an opposite cold and hot working state according to the signed residual heat state; and updating a temperature dynamic model and the signed residual heat state according to an actual process object temperature; the application overcomes the problem that existing temperature control technologies are prone to cause secondary overshoot and energy waste in the cold and hot switching process, and significantly improves the operation energy efficiency and control stability of an industrial temperature control system.

Description

Technical Field

[0001] This invention relates to the field of energy consumption optimization process control method for industrial cooling and heating temperature control units, and more particularly to an energy consumption optimization process control method and system for industrial cooling and heating temperature control units. Background Technology

[0002] This invention relates to the field of industrial temperature control technology, specifically to a method and system for optimizing the energy consumption of industrial cold and hot temperature control units for semiconductor manufacturing equipment process cold plates. With the continuous miniaturization of integrated circuit manufacturing processes and the increase in wafer size, core processes such as photolithography, etching, and thin film deposition place extremely stringent requirements on the temperature stability of the process environment. As a key thermal management unit, the process cold plate exchanges heat with the wafer stage or chamber walls through a circulating medium to achieve rapid and precise temperature regulation. Currently, a water-glycol mixture is commonly used as the circulating medium in such applications, as it possesses excellent thermal conductivity and a low freezing point, enabling it to adapt to wide temperature range operating requirements. A typical process usually includes four stages: pre-cooling, low-temperature holding, heating, and isothermal control. Significant temperature transitions need to be completed within a short time between each stage, such as rapidly cooling from room temperature to a low temperature and then quickly heating back to a high temperature. This poses a severe challenge to the response speed and energy efficiency of the temperature control system.

[0003] In practical engineering deployments, due to the limitations of internal space layout and cleanroom requirements of semiconductor manufacturing equipment, cooling and heating temperature control units are often placed in areas far from the process cold plates, connected by long supply and return liquid pipelines. This long-distance transport structure significantly increases the total volume of the water-glycol mixture contained in the circulation loop, far exceeding the volume held by the heat exchangers and process cold plates themselves. The large liquid volume means the system has a huge heat capacity. When the unit switches from cooling mode to heating mode (or vice versa) according to process instructions, a large amount of circulating medium remaining at the previous operating temperature in the pipeline will continuously flow into the process cold plates, forming so-called "residual cooling" or "residual heating" interference. For example, after the pre-cooling stage ends and the low-temperature holding stage begins, if a heating operation is immediately required, the remaining low-temperature liquid in the pipeline will severely slow down the heating rate, causing the actual process temperature to overshoot or deviate from the set value for an extended period. Similarly, when switching from high to low temperature, the remaining high-temperature liquid will delay the cooling process and increase the additional cooling load.

[0004] Traditional temperature control strategies are primarily based on proportional-integral-derivative (PID) regulation or its improved algorithms, which focus solely on the error between the current measured value and the target value, adjusting heater power or valve opening accordingly. However, when faced with the strong thermal inertia caused by large pipeline volumes and the residual energy impact during temperature reversals, relying solely on instantaneous deviation feedback is often insufficient. On one hand, PID controllers are sensitive to delays; transmission lag introduced by long pipelines can easily lead to system oscillations or even instability. On the other hand, the release of residual energy is a nonlinear process that decays over time. Conventional control laws cannot actively predict its impact amplitude and duration, and can only passively wait for the deviation to converge. This not only prolongs process changeover time but also results in significant wasted energy consumption—for example, in the initial stage of heating, the heater needs to consume additional electrical energy to offset the heat absorption effect of residual cold, and after the residual cold is exhausted, overshoot may require refrigeration intervention, creating an energy offset. Therefore, existing technologies struggle to simultaneously meet the dynamic performance requirements of rapid temperature response in semiconductor manufacturing and the increasingly stringent energy-saving and consumption-reducing targets for industrial temperature control systems.

[0005] In summary, there is an urgent need to develop a control method and system that can fully consider the thermal inertia characteristics of the system, effectively suppress residual energy disturbances, and achieve energy consumption optimization throughout the entire process, thereby improving the operating efficiency and temperature control quality of the cooling and heating temperature control unit in the application of semiconductor process cold plates. Summary of the Invention

[0006] This invention overcomes the shortcomings of the prior art and provides a method and system for optimizing the energy consumption process control of industrial cooling and heating temperature control units.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a process control method for optimizing energy consumption in industrial cooling and heating temperature control units, comprising: S1: Acquire process control data and process temperature program for industrial temperature control processes; The process control data includes the temperature of the process object, the supply temperature of the circulating medium, the return temperature, the circulation flow rate, the cooling input, and the heating input. S2: Determine the dynamic temperature tolerance range corresponding to the current process stage based on the process temperature program and process control data; S3: Determine the marked residual heat state, which represents the unreleased cold or heat, based on process control data; S4: Based on the signed residual heat state and temperature dynamics model, predict the future process object temperature and process energy consumption corresponding to the candidate process control sequence; With the constraint that the future process object temperature is within the dynamic temperature tolerance range, and with the goal of reducing process energy consumption and the cost of switching between hot and cold, rolling optimization is performed to obtain the target process control sequence. S5: When the target process control sequence indicates a reversal of the hot and cold operating states, determine the residual heat release process and the start-up conditions for the opposite hot and cold operating states based on the marked residual heat state. Process control is executed according to the target process control sequence, and the temperature dynamic model and the signed residual heat state are updated according to the actual process object temperature.

[0008] In a preferred embodiment of the present invention, in S2, the dynamic temperature tolerance range is determined in the following manner: Identify the current process stage and obtain the target temperature, allowable temperature deviation, allowable heating rate, allowable cooling rate, and remaining time of the stage; By combining the thermal inertia parameters of the process object, the upper and lower temperature boundaries at each moment in the prediction time domain are generated, thus obtaining the dynamic temperature tolerance domain.

[0009] In a preferred embodiment of the present invention, in S3, the marked residual heat state is determined by the following method: The heat exchange rate is determined based on the supply temperature, return temperature, and circulation flow rate. The heat balance is determined based on the cooling input, heating input, heat exchange, process heat load, and environmental heat exchange loss. The current residual heat state is updated based on the residual heat state, heat balance difference, and residual heat retention coefficient of the previous control cycle. Positive values ​​indicate residual heat effect, and negative values ​​indicate residual cold effect.

[0010] In a preferred embodiment of the present invention, in S4, the rolling optimization includes: Determine the cooling energy consumption, heating energy consumption, circulation transport energy consumption, temperature out-of-bounds amount, number of cooling and heating switching times, and control quantity change range corresponding to each candidate process control sequence; Under the conditions that cooling output and heating output are mutually exclusive, and the temperature change rate and circulation flow rate meet preset constraints, candidate process control sequences that meet preset optimization conditions in terms of comprehensive evaluation value are selected.

[0011] In a preferred embodiment of the present invention, in S5, the residual heat release process includes: Reduce the output corresponding to the current cooling and heating operating state and then stop the cooling and heating output; Adjust the circulation flow rate and flow path distribution based on the marked residual heat state and the temperature margin between the process object temperature and the corresponding temperature boundary. Predict the temperature change caused by residual heat and activate the opposite hot and cold working state when the start-up conditions are met.

[0012] In a preferred embodiment of the present invention, the activation conditions include at least: The absolute value of the signed residual heat state is not greater than the residual heat threshold; Predicting temperature changes will not cause the temperature of the process object to exceed the corresponding temperature boundary. Without initiating the opposite hot / cold operating state, the temperature of the process object cannot enter the dynamic temperature tolerance range within a preset time.

[0013] In a preferred embodiment of the present invention, in S5, the update includes: Obtain the temperature prediction deviation between the actual temperature of the process object and the predicted temperature of the process object; Based on temperature prediction deviations and process control data, update the equivalent heat capacity, heat transfer coefficient, transport hysteresis time, and residual heat retention coefficient. If the temperature prediction deviation exceeds the preset threshold, the rolling optimization is re-executed.

[0014] An energy consumption optimization process control system for an industrial cooling and heating temperature control unit includes: The data acquisition module is used to acquire process control data and process temperature programs; The generation module is used to determine the dynamic temperature tolerance range; The observation module is used to determine the state of the signed residual heat. The prediction module is used to predict the future temperature of the process object and the energy consumption of the process. The optimization module is used to obtain the target process control sequence under dynamic temperature tolerance domain constraints; The control module is used to determine the residual heat release process and the start-up conditions for the opposite hot and cold operating states; The output module is used to output process control commands corresponding to refrigeration, heating, circulation flow, and flow path allocation; The update module is used to update the temperature dynamics model and residual heat state.

[0015] In a preferred embodiment of the present invention, the acquisition module includes a process object temperature sensor, a supply liquid temperature sensor, a return liquid temperature sensor, a flow detector, and an energy input detector; the output module includes a cooling output control interface, a heating output control interface, a circulating flow control interface, and a flow path distribution control interface; each module is located in the process controller.

[0016] In a preferred embodiment of the present invention, an internal buffer flow path and a flow path switching mechanism are connected in parallel with the heat exchange flow path of the process object; the control module is used to control the flow path switching mechanism according to the marked residual heat state and temperature margin, and to prohibit the output of cooling or heating process control commands with opposite polarity before the start-up conditions are met.

[0017] This invention addresses the shortcomings of the prior art and has the following beneficial effects: (1) This invention provides an energy consumption optimization process control method and system for industrial cold and hot temperature control units. By acquiring process control data and process temperature program, under the constraint of dynamic temperature tolerance domain, based on the signed residual heat state, the future process object temperature and process energy consumption are predicted and rolled optimization is performed, realizing refined energy consumption control. By calculating the net energy in the circulation system and using it as the signed process state, the propagation effect of residual cold or heat in the pipeline after the cold and heat source stops is accurately identified. This changes the limitation of traditional temperature control that only relies on instantaneous temperature deviation for control. From the global perspective of the process cycle, under the premise of meeting temperature accuracy, it effectively reduces the energy loss caused by the ineffective start and stop and mutual cancellation of cold and hot actuators. It overcomes the problem that existing temperature control technology is prone to secondary overshoot and energy waste during the cold and hot switching process, and significantly improves the operating energy efficiency and control stability of industrial temperature control system.

[0018] (2) This invention provides an energy consumption optimization process control method and system for industrial cold and hot temperature control units. By constructing a dynamic temperature tolerance domain, the temperature control boundary is expanded from a fixed dead zone to a predictive space that changes over time. This enables the controller to actively utilize the thermal inertia of the process object, pipeline, and heat exchanger for energy regulation, thereby unloading the cold and heat sources in advance within the allowable deviation range. Partial temperature transition is completed through the residual energy of the circulating medium. Compared with the traditional control method that simply pursues tracking a fixed target temperature and frequently starts the compressor or heater, the three-stage polarity switching mechanism introduced in this invention actively plans and diverts the residual heat release process when the cold and hot working states reverse. This not only alleviates the severe heat flow impact during the cold and hot switching process, but also eliminates the reverse energy consumption caused by prematurely starting the opposite polarity mechanism, such as cooling before heating or heating before cooling. Under the condition of ensuring the temperature control accuracy and heating / cooling rate requirements of complex industrial processes, deep optimization of unit energy consumption is achieved.

[0019] (3) This invention provides an energy consumption optimization process control method and system for industrial cold and hot temperature control units. By integrating an online update mechanism for the temperature dynamic model and residual heat state into the control architecture, and by acquiring the actual process object temperature and prediction deviation in real time, the equivalent heat capacity, heat transfer coefficient and delivery lag time can be dynamically corrected according to the changes in operating parameters. This ensures the high adaptability of the optimized control model to the actual industrial production environment, thereby solving the problem of inconsistent model input caused by asynchronous multi-source data, different ranges and short-term data loss. The system can adapt to different pipeline lengths, circulation flow and process load changes, and overcomes the defect of traditional temperature control systems that cannot cope with complex operating condition fluctuations due to parameter fixation. It realizes low energy consumption and high following capability collaborative control of the temperature control unit throughout the entire operating cycle, and further enhances the robustness of the system in high-precision industrial application scenarios such as semiconductors, chemicals and precision manufacturing. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a flowchart of an energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to the present invention; Figure 2 This is a structural block diagram of an energy consumption optimization process control system for an industrial cooling and heating temperature control unit according to the present invention; Figure 3 The flowchart shows the residual heat release process and the method for determining the start-up conditions for the opposite hot and cold working states. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below. Example 1

[0023] like Figure 1 As shown, an energy consumption optimization process control method for an industrial cooling and heating temperature control unit includes the following steps: S1: Acquire process control data and process temperature program for industrial temperature control processes; wherein, process control data includes process object temperature, circulating medium supply temperature, return temperature, circulation flow rate, cooling input and heating input.

[0024] S2: Determine the dynamic temperature tolerance range corresponding to the current process stage based on the process temperature program and process control data.

[0025] S3: Determine the marked residual heat state, which represents the unreleased cold or heat, based on process control data.

[0026] S4: Based on the signed residual heat state and temperature dynamic model, predict the future process object temperature and process energy consumption corresponding to the candidate process control sequence; with the future process object temperature being within the dynamic temperature tolerance domain as a constraint, and with the goal of reducing process energy consumption and the cost of switching between hot and cold, perform rolling optimization to obtain the target process control sequence.

[0027] S5: When the target process control sequence indicates a reversal of the hot and cold working states, determine the residual heat release process and the start-up conditions for the opposite hot and cold working states based on the marked residual heat state; execute process control according to the target process control sequence, and update the temperature dynamic model and the marked residual heat state based on the actual process object temperature.

[0028] It should be noted that the present invention first uses the process procedure to determine the boundary of the allowable temperature change, then regards the unreleased cold or heat in the circulation system as a directional process state, and then compares the temperature trajectory and energy consumption generated by different cooling and heating modes, output power, circulation flow rate and split ratio in the prediction time domain. The controller unloads the cold and heat source in advance without violating the process constraints, uses the residual energy to complete part of the temperature regulation, and starts the opposite cold and heat source when the residual energy will not continue to cause the boundary to be exceeded or is no longer sufficient to complete the regulation.

[0029] This addresses the issue that in existing technologies, residual cold or heat cannot be directly measured by a single sensor, as its magnitude is influenced by a combination of factors, including refrigeration or heating input, circulation flow rate, supply and return liquid temperature difference, pipeline volume, heat exchanger heat capacity, process heat load, and environmental heat dissipation. Furthermore, the dynamic temperature tolerance range cannot be simply expanded, as this could compromise process accuracy. Additionally, energy consumption, temperature deviation, and mode switching in rolling optimization have different dimensions, and direct weighting of these factors can lead to distorted evaluations. Example 2

[0030] S1: Acquire process control data and process temperature program for industrial temperature control processes; The process control data includes the temperature of the process object, the supply temperature of the circulating medium, the return temperature, the circulation flow rate, the cooling input, and the heating input. It should be noted that the present invention takes an industrial cold and hot temperature control unit with a compression refrigeration circuit, an electric heating circuit, a variable frequency circulating pump, a three-way flow path switching mechanism and an internal buffer flow path as an example, and the process objects are semiconductor process cavity cold plates, chemical reactor jackets, power battery test cold plates, precision machine tool spindle cooling circuits or material cold and hot cycle test chambers.

[0031] This embodiment uses the process cold plate of semiconductor manufacturing equipment as the application object. The circulating medium is a water-glycol mixture. The process includes pre-cooling, low temperature holding, heating and constant temperature stages. In this application scenario, the pipeline connecting the unit and the process cold plate is relatively long and the circulating medium has a large volume. When the hot and cold states are reversed, the residual cold or residual heat has a significant impact on the subsequent process temperature.

[0032] The process control in this invention refers to the process of generating control commands based on the temperature of the process object and the state of the circulation loop, and changing the temperature of the process object through cooling output, heating output, circulation flow rate and flow path allocation. The process control data all come from the sensor and actuator feedback and process formula of the industrial temperature control system itself, and does not involve personal information.

[0033] In each control cycle, temperature and flow rate are read from detection points installed on the process object, unit outlet, unit return port, and circulation pipeline; current input quantities are read from the refrigeration and heating execution channels; and the currently effective process formula is read, thereby forming a data frame describing the current state of the industrial temperature control process.

[0034] Among them, the process object temperature refers to the temperature that reflects the actual thermal state of the process object, specifically the temperature of the surface of the cold plate, the material in the reactor, or the inner wall of the jacket; the supply liquid temperature is the temperature of the circulating medium before it leaves the unit and enters the heat exchange flow path of the process object; the return liquid temperature is the temperature of the circulating medium after it completes heat exchange and returns to the unit; the circulation flow rate is the volume or mass of the medium passing through the circulation loop per unit time; the cooling input and heating input respectively characterize the intensity of the current cooling and heating effects, expressed by electric power, compressor frequency, estimated cooling capacity, or normalized output percentage; the process temperature program is a process formula that specifies the target temperature, stage duration, allowable deviation, and heating / cooling rate according to time or event.

[0035] The acquired process control data serves as the common input for both the temperature dynamic model and the energy balance model. The supply and return liquid temperature difference and flow rate together reflect the heat flow transferred from the circulating medium to the process object; the cooling input and heating input reflect the amount of cold or heat injected into the circulating system by the unit; and the temperature change of the process object reflects the result of the above energy action. These quantities are synchronized in time to distinguish between temperature response lag caused by delivery delay and sensor abnormalities or changes in external load.

[0036] The specific operation involves: reading the raw sampled values ​​of the process object temperature, supply liquid temperature, return liquid temperature, and circulation flow rate; reading the feedback values ​​of the refrigeration output channel, heating output channel, circulation pump output channel, and flow path switching channel; reading the current stage number, target temperature, stage start time, and stage end condition in the process temperature program; aligning the data according to a unified timestamp; performing quality judgment on temperature spikes, flow spikes, and short-term communication packet loss; and combining the data that have completed the quality judgment into the process control data for the current control cycle.

[0037] The control cycle Δt is preferably 0.2-10S; for semiconductor process cold plates and battery testing cold plates, it is preferably 0.5-2S; for large inertial objects such as reactors, it can be 2-10S. The temperature sensor resolution is preferably no greater than 0.1℃, and the time deviation between the supply and return liquid temperature sensors is preferably no greater than one control cycle. For data missing for no more than 3 consecutive control cycles, the weighted result of the most recent effective value and linear extrapolation is used to fill the gaps. When data is missing for more than 3 consecutive control cycles, the rolling optimization is paused and the preset safety control strategy is switched.

[0038] To suppress high-frequency noise from the sensor, a first-order low-pass filter or a sliding median filter with 3 to 9 sampling points is used for the raw temperature. The filtering time constant should be less than one-tenth of the main time constant of the controlled object to avoid introducing significant lag through the filtering itself. Outliers are judged by both the rate of change of adjacent samples and the physical range. When the rate of temperature change exceeds 1.5 to 3 times the maximum allowable heating and cooling rate of the equipment and no change in the same direction is observed at adjacent redundant measuring points, the sample is marked as an outlier. This solves the problem of inconsistent model input caused by asynchronous multi-source data, different measurement ranges, and short-term missing data, and provides a common time reference for subsequent calculations of dynamic temperature tolerance domain and residual heat state.

[0039] S2: Determine the dynamic temperature tolerance range corresponding to the current process stage based on the process temperature program and process control data; The dynamic temperature tolerance domain is determined by: identifying the current process stage, obtaining the target temperature, allowable temperature deviation, allowable heating rate, allowable cooling rate, and remaining time of the stage; and combining the thermal inertia parameters of the process object to generate the upper and lower temperature boundaries at each moment in the prediction time domain, thereby obtaining the dynamic temperature tolerance domain.

[0040] It should be noted that in this step, the controller first determines whether it is currently in the heating, constant temperature, cooling, or hot-cold transition stage. Then, based on the temperature accuracy and heating / cooling rate specified in the process formula, combined with the current thermal inertia of the process object and the remaining stage time, it calculates the maximum and minimum allowable temperatures for each predicted moment in the future.

[0041] The dynamic temperature tolerance domain is a time series consisting of the upper and lower temperature boundaries. Unlike the dead zone, which sets fixed positive and negative deviations around the setpoint, the boundary can change with time in the prediction time domain. In the isothermal stage, the boundary is mainly determined by the allowable temperature deviation and temperature uniformity requirements. In the heating or cooling stage, the boundary is also constrained by the allowable heating / cooling rate and the remaining time of the stage. In the thermal conversion stage, the boundary is further combined with the direction of residual heat and model uncertainty to set a safety margin.

[0042] The process temperature program specifies the acceptable temperature range for a product or process. When the process object has a large heat capacity, the output of the cold and heat sources is reduced in advance within the allowable deviation range. The existing heat is used to continue heating or cooling, which significantly reduces peak power and ineffective compensation. The dynamic temperature tolerance domain makes this process allowable space explicit and ensures that the energy-saving trajectory does not impair process requirements through the upper and lower boundaries.

[0043] The specific operation is as follows: determine the current process stage based on the stage identifier of the process temperature program and the current time; read the target temperature, allowable temperature deviation, allowable heating rate, allowable cooling rate and remaining time of the stage; estimate the equivalent heat capacity and transport hysteresis based on the heat flow and temperature changes of the most recent control cycles; normalize the heat capacity, transport hysteresis and target temperature change rate; calculate the tolerance half-width at each moment in the prediction time domain; and superimpose the temperature change rate constraint and safety margin on the tolerance half-width to form the upper temperature boundary and lower temperature boundary.

[0044] To avoid dimensional confusion caused by the direct participation of different physical quantities in weighting, the following normalization method is adopted for variables such as heat capacity, transport hysteresis, temperature change rate, and energy consumption: .

[0045] Where, x j x is the variable to be normalized. j,min and x j,max These represent the lower and upper limits of the variable under the permissible operating conditions of the equipment or in the identified samples, respectively. ε is a positive number to prevent the denominator from being zero, and is preferably taken as 10. -6 Up to 10 -3 The normalized results are used to evaluate the function and tolerance width calculations, but do not replace the dimensional variables in the physical model.

[0046] In the above, the tolerance half-width is determined according to the following formula: .

[0047] Where, δ i δ0 represents the tolerance half-width for predicting the i-th time step in the time domain; δ0 represents the basic allowable deviation specified in the process procedure. To normalize the equivalent heat capacity; Normalized transport delay; λ is the normalized target temperature change rate; C , λ τ and λ R The dimensionless adjustment coefficient; clip indicates that the calculation result is restricted to δ. min With δ max Between these, the greater the equivalent heat capacity and transport hysteresis, the greater the space that the controller can unload in advance; the faster the target temperature changes, the more the tolerance half-width should be narrowed to ensure the tracking capability.

[0048] The upper and lower temperature boundaries are determined based on the tolerance half-width: .

[0049] Among them, T ref,k+i The reference temperature for the process temperature program at the predicted time k+i will also be T during the heating phase. U,k+i The temperature is limited to no higher than the upper limit formed by the current temperature and the allowable heating rate; for the cooling phase, T... L,k+i The limit is set to be no lower than the lower limit formed by the current temperature and the allowable cooling rate. If the remaining time of the stage is less than the prediction time domain, the target temperature and allowable deviation of the next process stage are used in the latter part of the prediction time domain to avoid sudden changes in control direction at the stage boundary.

[0050] Prediction time domain length N p Preferably, the number of control cycles is 30 to 300, and the total predicted duration should not be less than the smaller of the circulating medium transport hysteresis and the main time constant of the process object; δ min The value can be 1.5 to 3 times the overall error of the temperature sensor; δ max The temperature difference should not exceed the allowable temperature difference specified in the process formula; λ C , λ τ and λ R Data fitting can be achieved through equipment debugging, with an optimal range of 0.05 to 0.8. If the model prediction error increases, the safety margin can be increased or δ can be reduced. i Tighten the tolerance range.

[0051] This establishes a calculable constraint boundary between temperature accuracy and energy consumption, and dynamically allocates allowable deviations based on process stage and thermal inertia, enabling subsequent rolling optimization to select a low-energy trajectory that is physically feasible and meets process requirements.

[0052] S3: Determine the marked residual heat state, which represents the unreleased cold or heat, based on process control data; The marked residual heat state is determined as follows: the heat exchange is determined based on the supply liquid temperature, return liquid temperature and circulation flow rate; the heat balance is determined based on the cooling input, heating input, heat exchange, process heat load and environmental heat exchange loss; the current residual heat state is updated based on the residual heat state, heat balance and residual heat retention coefficient of the previous control cycle, where a positive value indicates residual heat effect and a negative value indicates residual cooling effect.

[0053] It should be noted that this step calculates the net energy that has not yet been released in the circulation system based on the cooling or heating capacity already input into the circulation system, the actual heat transferred by the circulation medium to the process object, and the heat lost or absorbed by the pipeline to the environment. Positive and negative signs are used to distinguish whether it will lead to a subsequent temperature rise or fall.

[0054] The unit for the signed residual heat state is joules. A positive value indicates that the circulation system still has the ability to continue to raise the temperature of the process object even if the current heat source stops immediately; a negative value indicates that the circulation system still has the ability to continue to lower the temperature of the process object. This state is not the absolute value of the heat in a container, but the net heat release effect relative to the current temperature process and reference state.

[0055] When the supply temperature of the circulating medium is higher than the return temperature, the circulating medium releases heat to the process object; when the supply temperature is lower than the return temperature, the circulating medium absorbs heat from the process object. The temperature difference between the supply and return liquids multiplied by the density, specific heat capacity, and volumetric flow rate yields the directional heat exchange power. By integrating the unit's heat input, heat exchange output to the process object, and environmental losses over time, the remaining energy inside the circulating system can be estimated.

[0056] The signed heat transfer power transferred from the circulating medium to the process object is calculated according to the following formula: .

[0057] in, ρ is the heat exchange power in the kth control cycle, in W; ρ is the density of the circulating medium, in kg / m³. 3 ;c p q is the specific heat capacity at constant pressure, expressed in J / (kg·K); v,k Volumetric flow rate, unit: m³ 3 / s;T s,k and T r,k These are the supply temperature and the return temperature, respectively, according to the notation conventions of this embodiment. A value greater than zero indicates that the circulating medium supplies heat to the process object, while a value less than zero indicates that the circulating medium absorbs heat from the process object.

[0058] The signed residual heat state for the current control cycle is updated according to the following formula: .

[0059] Among them, E r,k and E r,k-1 These are the residual heat states for the current and previous control cycles, respectively; α k η is the residual heat retention coefficient. h and η c These are the coefficients for converting heating input and cooling input into effective heat flux in the circulating system, respectively; P h,k and P c,k These are the heating input and the cooling input, respectively. The value represents the signed heat transfer loss between the circulating loop and the environment. All terms in the formula are in units of W, and after multiplying by the control period Δt, the unit is joule. Therefore, the dimensions of the left and right sides of the formula are consistent.

[0060] The residual heat retention factor is used to describe the energy decay of a circulating system over time, and its preferred expression is: .

[0061] Where, τ r,k The residual heat decay time constant is determined online based on the circulating medium volume, pipeline insulation performance, heat exchanger heat capacity, and circulating flow rate. This exponential form ensures that 0 < α. k ≤1; τ r,k The larger the value, the longer the energy remains in the circulatory system.

[0062] The specific operation involves determining the density and specific heat capacity based on the type of circulating medium and the current average temperature; calculating the signed heat transfer power based on the supply liquid temperature, return liquid temperature, and circulation flow rate; determining the effective input power of the cold and heat sources based on the cooling input and heating input and the corresponding conversion coefficients; estimating the ambient heat transfer loss based on the ambient temperature, pipe surface area, and equivalent heat transfer coefficient; updating the current residual heat state based on the residual heat state of the previous cycle and the current heat balance; and correcting the residual heat retention coefficient based on the actual temperature response.

[0063] Environmental heat transfer loss adopts Calculate, where, This is the equivalent heat transfer coefficient of the piping and heat exchanger to the environment. For equivalent heat transfer area, The average temperature of the supply and return liquids. The ambient temperature. and It can be identified based on the decay curve of the average medium temperature in a test where the cold and heat sources are turned off and the circulating pump maintains a constant flow rate.

[0064] The density and specific heat capacity of the water-ethylene glycol mixture are preferably obtained by referring to tables based on the medium concentration and temperature, or through equipment calibration. η h The preferred value is 0.85 to 1.05; η cIt's not the compressor's energy efficiency ratio itself, but rather the calibration coefficient from the refrigeration input to the effective refrigeration power on the circulation side. If P c Expressed in terms of electrical power, then η c Take values ​​from 1.5 to 5.0, if P c Since cooling capacity has already been used as the representation, then η c The value can be taken as 0.85 to 1.15, τ r,k The preferred duration is 5S to 600S, with the specific range determined by the cycle volume and insulation structure.

[0065] By using the marked residual heat state, it can be determined whether the current temperature deviation will decrease on its own or continue to increase without any opposite hot or cold output. For example, if the process object temperature is lower than the target temperature but E r,k When E is still negative, it indicates that the circulation system still has a cooling effect; in this case, directly starting heating will result in energy cancellation. Conversely, if E r,k If the temperature is close to zero and the predicted temperature cannot enter the tolerance range, heating needs to be started promptly.

[0066] S4: Based on the signed residual heat state and temperature dynamics model, predict the future process object temperature and process energy consumption corresponding to the candidate process control sequence; With the constraint that the future process object temperature is within the dynamic temperature tolerance range, and with the goal of reducing process energy consumption and the cost of switching between hot and cold, rolling optimization is performed to obtain the target process control sequence. Rolling optimization includes: determining the cooling energy consumption, heating energy consumption, circulating transport energy consumption, temperature out-of-bounds amount, number of cooling and heating switching times, and control quantity change range corresponding to each candidate process control sequence; and selecting candidate process control sequences whose comprehensive evaluation values ​​meet the preset optimization conditions under the conditions that cooling output and heating output are mutually exclusive and the temperature change rate and circulating flow rate meet the preset constraints.

[0067] It should be noted that in this step, several future control schemes are generated by the controller, and the temperature of the process object, the energy required by the equipment, and the number of times the hot and cold modes will switch are simulated under these schemes. The scheme with the lowest overall cost is selected from the schemes that meet the temperature boundary and equipment constraints, and only its first control cycle is executed. The scheme is then recalculated using new data in the next cycle.

[0068] The candidate process control sequence consists of the cooling and heating operating states, cooling output, heating output, circulation flow rate, and flow path allocation in the prediction time domain, where the cooling and heating operating states are m. i -1, 0, and 1 represent the cooling state, the state of residual heat release or retention without cold or heat source output, and the heating state, respectively. The flow path distribution can be based on the flow rate ratio γ entering the heat exchange path of the process object. i This means that 0 ≤ γ i ≤1, the remaining flow enters the internal buffer flow path.

[0069] Temperature prediction employs a heat dissipation equilibrium model with explicit physical meaning to describe the dynamic lag between the output of the heat source and the actual heat exchange power. A first-order inertial element for the heat exchange power is first established: .

[0070] Where, τ q The time constant is used to establish the heat exchange power output from the cold and heat sources to the circulation side. The cooling input acts as a negative heat flow in the formula, while the heating input acts as a positive heat flow. If an internal buffer flow path is used, the effective heat exchange power entering the process object is γ. i The portion entering the buffer flow path does not directly act on the process object.

[0071] The temperature of the process object is predicted according to the following formula: .

[0072] Among them, T p,k+i To predict the temperature of the process object; C eq d represents the equivalent heat capacity of the process object and its directly heat-exchanging parts; d represents the transport hysteresis expressed in control cycle number. For process load disturbance, a positive value indicates heat input to the process object; hA is the product of the equivalent heat transfer coefficient and area between the process object and the environment; T a The value in parentheses in this formula is the ambient temperature. Multiplying it by Δt and dividing by J / (kg·K) yields the Kelvin, which has the same dimensions.

[0073] The process load disturbance is obtained by back-calculation based on the heat balance of historical control cycles, and extended in the prediction time domain using the most recent effective value, linear trend or stage correlation average. For periodic process equipment, the average load curve of the previous several batches of the same process formula in the same stage is used as the prior value.

[0074] The rolling optimization described above specifically involves generating candidate cold and hot operating state sequences based on the current process stage, residual heat status, and equipment allowable modes; discretely or continuously optimizing the cooling output, heating output, circulation flow rate, and flow path allocation under each operating state; progressively predicting the future temperature and heat exchange power of the process object using a temperature dynamic model; calculating the cooling energy consumption, heating energy consumption, circulation transport energy consumption, temperature out-of-bounds amount, number of cold and hot switching times, and control quantity change amplitude for each candidate sequence; eliminating sequences that violate mutual exclusion constraints, temperature change rate constraints, flow rate constraints, and minimum running time constraints; and selecting the target process control sequence whose comprehensive evaluation value meets the preset optimization conditions.

[0075] The comprehensive evaluation function can be set as follows: .

[0076] Where, N p To predict the length of the time domain; Normalized process energy consumption, including energy consumption for cooling, heating, and circulating transport; φ i Penalty for exceeding temperature limits; m i Operating in both hot and cold states; i w is the normalized control vector. E w T w M and w U The weights are for energy consumption, temperature exceedance, mode switching, and control changes. After normalization, each evaluation item is dimensionless to avoid the direct addition of different dimensions.

[0077] Temperature violation penalty is calculated according to the following formula: .

[0078] Where, ΔT b As a temperature normalization reference, it is preferably taken as the process allowable temperature difference or 1℃, when the temperature is within the dynamic temperature tolerance range. i The penalty is zero; when the temperature is above the upper boundary or below the lower boundary, the penalty increases with the square of the out-of-bounds amount, thus allowing the optimizer to make selections based primarily on energy consumption and switching costs within the tolerance domain, while maintaining strong suppression of temperature out-of-bounds.

[0079] Cooling energy consumption can be expressed in P (units per liter). c ,iΔt / COP i Calculate, where COP i The heating energy consumption is obtained based on the supply and return liquid temperatures, ambient temperature, and compressor operating frequency; it can be calculated as P. h ,iΔt / η h Calculation; the energy consumption of the circulating transport can be determined based on the pump power curve P. p =a0+a1q v +a2q v² +a3q v³ If the pump power has been directly acquired by the energy input detector, then the measured power can be used to establish a piecewise linear mapping between the flow rate and the pump power.

[0080] The constraints must include at least: P c i×P h i=0, to prevent simultaneous cooling and heating outputs; 0≤P c,i ≤P c,max ; 0≤P h,i ≤P h,max ;q v,min ≤q v,i ≤q v,max ;0≤γ i≤1; the predicted temperature of the process object does not exceed the upper and lower limits of equipment safety; the rate of temperature change does not exceed the allowable value of the process procedure; the continuous operation and shutdown time of the compressor meet the equipment protection requirements.

[0081] Control time domain N c It can be less than or equal to the prediction time domain N p Preferably, there are 5 to 60 control cycles; refrigeration, heating, and flow can each be represented by 3 to 15 discrete levels to form candidate sequences, or they can be solved using sequential quadratic programming, mixed integer programming, or heuristic search. T The value should be set such that temperature safety constraints take precedence over energy-saving targets, preferably W. E 5 to 100 times; w M The preferred option is w E 0.2 to 10 times; w U The preferred option is w E The actual weight is calibrated by the allowable temperature difference of the equipment and the energy loss per switching, ranging from 0.01 to 2 times.

[0082] Rolling optimization executes only the first control action of the target process control sequence. The data is reacquired and solved in the next control cycle. This mechanism can continuously correct the future trajectory as the process load and model parameters change, avoiding long-term deviation of a one-time plan due to prediction errors.

[0083] S5: When the target process control sequence indicates a reversal of the hot and cold operating states, determine the residual heat release process and the start-up conditions for the opposite hot and cold operating states based on the marked residual heat state. Process control is executed according to the target process control sequence, and the temperature dynamic model and the signed residual heat state are updated according to the actual process object temperature.

[0084] The residual heat release process includes: reducing the output corresponding to the current hot and cold working state and then stopping the cooling output and heating output; adjusting the circulation flow rate and flow path distribution according to the marked residual heat state and the temperature margin between the process object temperature and the corresponding temperature boundary; predicting the temperature change caused by the residual heat and starting the opposite hot and cold working state when the start-up conditions are met.

[0085] The activation conditions include at least the following: the absolute value of the signed residual heat state is not greater than the residual heat threshold; the predicted temperature change will not cause the temperature of the process object to exceed the corresponding temperature boundary; and the temperature of the process object cannot enter the dynamic temperature tolerance range within a preset time when the opposite hot and cold working state is not activated.

[0086] It should be noted that in this step, when the optimization result requires switching from cooling to heating or from heating to cooling, the opposite cold or heat source is not immediately turned on. Instead, the current output is gradually reduced and the cold or heat source is turned off. Then, the direction and rate of release of the remaining cold or heat are controlled by the circulation pump and flow path switching mechanism. When the residual effect is small enough, or when waiting will not meet the temperature recovery requirements, the opposite cold or heat source is then turned on.

[0087] Even after the heat source and cold source are shut down, the residual heat in the circulation loop will still be transferred to the process object through the circulation medium. Its potential impact on the temperature of the process object can be estimated based on residual energy, equivalent heat capacity, transport hysteresis, and energy transfer efficiency. .

[0088] Where, ΔT r,k Temperature change of the process object caused by residual heat; k r E is the transfer coefficient of residual energy to the effective heat of the process object. r,k The sign represents the residual heat state; C eq τ is the equivalent heat capacity; d For transport lag; τ th Let ΔT be the thermal decay time constant. r,k With E r,k With the same sign, a positive value indicates continued heating, while a negative value indicates continued cooling.

[0089] The temperature margin between the current process object temperature and the corresponding temperature boundary is denoted as ΔT. margin,k If the residual effect is directed towards the upper boundary, then the temperature margin is T. U,k -T p,k If it faces the lower boundary, the temperature margin is T. p,k -T L,k The residual heat threshold, considering model error and sensor error, can be determined using the following formula: .

[0090] Where, ΔT safe For safety margin, if |E r,k | Less than or equal to E th,k This means that even if the remaining energy continues to act, the predicted temperature change will usually not exceed the corresponding boundary.

[0091] Specifically, the residual heat release and polarity switching are as follows: when the target process control sequence indicates a reversal of the cold and hot operating states, the current cooling output or heating output is reduced according to the preset maximum output change rate; after the current output is reduced to the stop threshold, the cooling output and heating output are shut off, while the circulating pump continues to run; the proportion of the circulating flow rate to the heat exchange path of the process object is determined based on the sign, absolute value, and temperature margin of the residual heat; the temperature change caused by the residual heat and the ability of the process object to enter the dynamic temperature tolerance range within a preset time are predicted; the release of residual heat is continued or the opposite cold and hot operating state is started based on the start-up conditions; and corresponding cooling, heating, circulating flow rate, and flow path allocation control commands are output.

[0092] When the direction of residual heat is favorable to the current process objective and will not cause temperature exceedance, the flow split ratio γ entering the heat exchange path of the process object can be increased, allowing the remaining energy to continue to act on the process object. When residual heat may lead to temperature exceedance, γ should be reduced and more circulating medium should enter the internal buffer flow path to reduce the actual heat flow received by the process object. When the residual heat is close to zero, γ can be restored to the normal heat exchange ratio. The internal buffer flow path can be composed of at least one of a buffer tank, a bypass pipe, and a heat exchanger, and its function is to allow residual energy that is not suitable for direct application to the process object to circulate, mix, or attenuate within the unit.

[0093] The startup conditions must consider at least the following three factors: First, whether the absolute value of the signed residual heat state is not greater than the residual heat threshold; second, whether the predicted temperature change will not cause the process object temperature to exceed the corresponding temperature boundary; and third, whether the process object temperature cannot enter the dynamic temperature tolerance range within a preset time when the opposite heating / cooling operating state is not started. Preferably, the opposite heating / cooling state is allowed to start normally when both the first and second factors are met; when the third factor is met, even if the residual heat has not completely decayed, startup can be initiated earlier under the condition of limiting the initial output slope and maintaining the safety constraint of the second factor, in order to prevent excessively long response times.

[0094] In the above, the original polarity unloading time is preferably 2 to 120 seconds; the rate of change of cooling or heating output is preferably limited to 0.5% to 20% of the rated output per second; the circulation flow rate during residual heat release is preferably 20% to 100% of the rated flow rate; and the flow split ratio γ entering the heat exchange path of the process object is preferably 0 to 1. The preset time can be determined based on the remaining time of the current stage and the allowable heating and cooling rates, and is preferably 10 to 300 seconds. Safety margin ΔT safe The preferred value is 1 to 3 times the sum of the overall error of the temperature sensor and the root mean square error of the model.

[0095] After the process control command is issued, the update specifically involves obtaining the temperature prediction deviation between the actual process object temperature and the predicted process object temperature; updating the equivalent heat capacity, heat transfer coefficient, transport hysteresis time, and residual heat retention coefficient based on the deviation and process control data; and shortening the effective time of the current target control sequence and re-executing rolling optimization when the temperature prediction deviation exceeds the preset threshold.

[0096] The model is updated using a recursive least squares approach. Let θ be the vector of parameters to be identified, and φ be the vector of parameters to be identified. k Let y be the regression vector consisting of the previous cycle temperature, heat exchange power, ambient temperature, and delayed input. k Given the current actual temperature of the process object, the parameters are updated as follows: .

[0097] Among them, K k This is the gain vector.

[0098] .

[0099] Among them, P k-1 Let λ be the parametric covariance matrix. f The forgetting factor is preferably between 0.95 and 0.999. A larger forgetting factor is suitable for stable equipment with slowly changing parameters, while a smaller forgetting factor is suitable for equipment with frequent changes in load or circulating medium. To prevent parameter drift, the equivalent heat capacity, heat transfer coefficient, and hysteresis time are limited to the physical feasibility of the equipment.

[0100] This allows this step to overcome the conventional bias that delaying the opposite heat source will inevitably reduce the response speed. For a circulating system with significant residual heat, immediately starting the opposite heat source will first offset the residual effect, and then the system will adjust in the opposite direction again due to the secondary temperature overshoot, which will increase the total settling time. By selecting the release timing according to the residual heat status, energy consumption, switching frequency and secondary overshoot can be reduced simultaneously under some operating conditions. Example 3

[0101] Existing temperature control solutions all treat thermal inertia as an adverse factor causing control lag. Therefore, they compensate for the lag by increasing the control gain or activating the opposite heat source in advance. However, this invention treats the predictable and releaseable part of thermal inertia as virtual thermal energy storage that can participate in temperature regulation. It determines the direction and quantity of remaining energy by the state of the marked residual heat, then determines how much remaining energy is allowed to be used in the process by the dynamic temperature tolerance domain, and finally determines whether the remaining energy should continue to act on the process object or enter the internal buffer flow path by rolling optimization and diversion control.

[0102] This establishes a connection between the dynamic temperature tolerance domain, the signed residual heat state, and the hot / cold polarity switching control. If only the dynamic temperature tolerance domain is set without observing the residual heat, the controller cannot determine how much the temperature will change after the hot / cold source is turned off in advance, which can easily lead to exceeding the limits. If only the residual heat is observed while still tracking a fixed temperature point, the controller lacks the allowable space to utilize the residual energy. If only a residual heat release stage is added without rolling prediction, the release direction and duration cannot be determined based on future process stages. This invention combines these three elements, allowing the controller to obtain the process allowable space, the internal energy state, and the computational mechanism for selecting the optimal control sequence.

[0103] This invention selects the supply liquid temperature, return liquid temperature, and circulation flow rate as the core parameters for residual heat calculation because these three can directly form the signed heat transfer power on the circulation side. Compared with estimating energy using only the compressor frequency or heating output, the supply and return liquid heat transfer power already includes the comprehensive effects of refrigeration circuit efficiency, heat exchanger dynamics, and actual flow rate changes. Compared with estimating using only the temperature change of the process object, it responds earlier and does not need to wait for a significant temperature change in a large heat capacity object. Therefore, this parameter combination can identify the cold or heat still propagating in the pipeline in the early stages after the cold or heat source stops.

[0104] This invention selects the equivalent heat capacity C eq The reason for using only the quality of the process object is that the actual heat capacity involved in short-term temperature changes also includes the jacket, cold plate, connectors, and part of the circulating medium. eq This can be obtained by applying a known step heat flux at a steady flow rate and measuring the initial temperature rise rate, i.e. When identifying the data, it is preferable to select the 20-300s range where the temperature change is approximately linear, and exclude data from valve switching and sudden load changes.

[0105] Transport hysteresis τ d The peak value of the cross-correlation between the change in the supply liquid temperature and the change in the process object temperature is obtained by changing the unit outlet temperature, or estimated by dividing the effective pipeline volume by the volumetric flow rate. The former can simultaneously reflect the sensor position and the mixing effect, while the latter is suitable for providing initial values ​​during the initial start-up of the equipment. During online operation, the τ value can be adjusted based on the prediction error. d Make minor adjustments.

[0106] In an exemplary industrial cold plate configuration, the circulating medium volume is 60L to 180L, the rated flow rate is 20-80L / min, and the equivalent heat capacity of the process object is 1×10⁻⁶. 5 J / (kg·K) up to 8×10 6J / (kg·K), delivery hysteresis 5S to 60S, rated cooling capacity 3KW to 30KW, rated heating capacity 2KW to 20KW. Control cycle is 1S, prediction time domain is 120S, control time domain is 30S, basic allowable temperature difference δ0 is 0.3℃, and maximum tolerance half-width δ max The temperature is set at 0.8℃, with a safety margin of 0.15℃. All of these parameters are within the commonly achievable range of industrial temperature control equipment.

[0107] In this configuration, when the process switches from the 10°C low-temperature holding stage to the 25°C heating stage, if the residual heat state of the circulation loop is negative, the refrigeration is first turned off and the circulation flow rate is maintained at 30% to 60% of the rated flow rate. At the same time, the flow ratio entering the heat exchange path of the process object is reduced to 0.2 to 0.5. As the residual cooling capacity decays, if the predicted temperature change will not cross the lower limit and the temperature cannot enter the tolerance range within a preset time without heating, the heating output is gradually increased. This process ensures that the heating power does not have to be used to offset all the residual cooling capacity first.

[0108] When the process switches from the high-temperature stage to the cooling stage and the residual heat is positive, symmetrical control logic is adopted. If the positive residual heat continues to act and will cause the temperature of the process object to exceed the upper boundary, the flow diversion ratio of the internal buffer flow path is increased. If the direction of the residual heat is still conducive to completing the current heating tail section, it is allowed to continue to act and the heating output is reduced in advance. Thus, the present invention does not bypass the residual heat in a fixed way, but selects the flow diversion ratio according to the direction of the residual heat, the temperature margin and the future stage. Example 4

[0109] The present invention also provides an energy consumption optimization process control system for industrial cooling and heating temperature control units, which includes a data acquisition module, a generation module, an observation module, a prediction module, an optimization module, a control module, an output module, and an update module.

[0110] The acquisition module is used to acquire process control data and process temperature programs. The acquisition module includes signal acquisition interfaces that connect to field process object temperature sensors, liquid supply temperature sensors, liquid return temperature sensors, flow detectors, and energy input detectors. Field sensors can be resistance temperature detectors, thermocouples, mass flow meters, volumetric flow meters, and power transmitters. The acquisition module performs timestamp alignment, filtering, anomaly detection, and missing value processing on analog or bus data.

[0111] The generation module is used to identify the current process stage and generate the upper and lower temperature boundaries in the prediction time domain based on the target temperature, allowable deviation, allowable heating rate, allowable cooling rate, remaining time of the stage, and thermal inertia parameters of the process object. The generation module provides the boundaries to the prediction module and the optimization module.

[0112] The observation module is used to update the residual heat state with symbols according to the energy balance based on the supply and return liquid temperature difference, circulation flow rate, cooling input, heating input and environmental heat exchange loss. The observation module is also used to correct the residual heat retention coefficient based on the actual temperature response.

[0113] The prediction module is used to output the process object temperature, heat exchange power and process energy consumption at each prediction time based on the temperature dynamic model, the signed residual heat state and the candidate process control sequence. The input of the prediction module is connected to the data output of the acquisition module, the generation module and the observation module respectively.

[0114] The optimization module is used to generate candidate process control sequences, calculate comprehensive evaluation values ​​and screen out sequences that violate constraints, and determine the sequences that meet the preset optimization conditions as target process control sequences. The optimization module provides the control module with the cold and hot working status plan and provides the output module with the cooling, heating, flow rate and flow path allocation settings for the current control cycle.

[0115] The control module is used to perform original polarity unloading, residual heat release and opposite polarity start determination when the target process control sequence indicates a reversal of the cold and hot states. The control module is connected to the control interface of the flow path switching mechanism and can send a polarity interlock signal to the output module when the start conditions are not met, prohibiting the output of cooling or heating commands that conflict with the current residual heat direction.

[0116] The output module includes a refrigeration output control interface, a heating output control interface, a circulation flow control interface, and a flow path distribution control interface. The refrigeration output control interface can be connected to a compressor inverter, an electronic expansion valve, or a refrigeration capacity regulator; the heating output control interface can be connected to a thyristor power regulator or a staged heating contactor; the circulation flow control interface is connected to a circulation pump inverter; and the flow path distribution control interface is connected to a three-way proportional valve, a bypass valve, or multiple switching valves.

[0117] The update module compares the actual temperature of the process object with the predicted temperature of the process object, updates the temperature dynamic model and residual heat state parameters based on the prediction deviation, and triggers the optimization module to solve the problem again when the deviation exceeds a preset threshold.

[0118] In terms of hardware structure, the functional logic of the generation module, observation module, prediction module, optimization module, control module, and update module can be set in the same process controller, or it can be implemented by an industrial computer and a programmable logic controller. The signal conditioning and communication interface of the acquisition module is set in the process controller, and the sensor body is arranged at the corresponding measurement point. The control interface of the output module is set in the process controller, and the actuator is arranged in the cold and hot temperature control unit and flow path. The modules transmit data to each other through a shared storage area, internal bus, or industrial Ethernet.

[0119] The heat exchange flow path and the internal buffer flow path of the process object are arranged in parallel. The flow path switching mechanism is set at the branching node or the merging node. By adjusting the flow ratio of the two flow paths, the effect of residual heat on the process object is changed. The internal buffer flow path preferably includes a buffer tank with a certain volume to mix the circulating media at different temperatures. It may also include a decay heat exchanger that exchanges heat with the environment or auxiliary cooling medium. The internal buffer flow path does not require long-term energy storage. Its main function is to temporarily reduce the proportion of residual energy that directly acts on the process object during the switching of hot and cold polarities.

[0120] When the system is running, the acquisition module generates process control data, the generation module outputs the dynamic temperature tolerance domain, the observation module outputs the signed residual heat state, the prediction module and the optimization module form the target process control sequence, the control module interlocks the target sequence according to the switching conditions, the output module drives the field actuators, and the update module corrects the parameters according to the actual response, thus forming a complete closed loop.

[0121] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A method for optimizing the process control of energy consumption in an industrial cooling and heating temperature control unit, characterized in that, include: S1: Acquire process control data and process temperature program for industrial temperature control processes; The process control data includes the temperature of the process object, the supply temperature of the circulating medium, the return temperature, the circulation flow rate, the cooling input, and the heating input. S2: Determine the dynamic temperature tolerance range corresponding to the current process stage based on the process temperature program and process control data; S3: Determine the marked residual heat state, which represents the unreleased cold or heat, based on process control data; Among them, the heating input is a positive energy input that increases the state value, and the cooling input is a negative energy input that decreases the state value; when the state value is positive, it means that the circulation system still has a residual heat effect that can continue to increase the temperature of the process object, and when it is negative, it means that the circulation system still has a residual cooling effect that can continue to decrease the temperature of the process object. S4: Based on the signed residual heat state and temperature dynamics model, predict the future process object temperature and process energy consumption corresponding to the candidate process control sequence; With the constraint that the future process object temperature is within the dynamic temperature tolerance range, and with the goal of reducing process energy consumption and the cost of switching between hot and cold, rolling optimization is performed to obtain the target process control sequence. S5: When the target process control sequence indicates a reversal of the hot and cold operating states, determine the residual heat release process and the start-up conditions for the opposite hot and cold operating states based on the marked residual heat state. Process control is executed according to the target process control sequence, and the temperature dynamic model and the signed residual heat state are updated according to the actual process object temperature.

2. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 1, characterized in that, In S2, the dynamic temperature tolerance range is determined as follows: Identify the current process stage and obtain the target temperature, allowable temperature deviation, allowable heating rate, allowable cooling rate, and remaining time of the stage; By combining the thermal inertia parameters of the process object, the upper and lower temperature boundaries at each moment in the prediction time domain are generated, thus obtaining the dynamic temperature tolerance domain.

3. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 1, characterized in that, In S3, the determination of the marked residual heat state is as follows: The heat exchange rate is determined based on the supply temperature, return temperature, and circulation flow rate. The heat balance is determined based on the cooling input, heating input, heat exchange, process heat load, and environmental heat exchange loss. The current residual heat state is updated based on the residual heat state of the previous control cycle, the heat balance difference, and the residual heat retention coefficient, where a positive value indicates residual heat effect and a negative value indicates residual cold effect.

4. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 1, characterized in that, In S4, the scrolling optimization includes: Determine the cooling energy consumption, heating energy consumption, circulation transport energy consumption, temperature out-of-bounds amount, number of cooling and heating switching times, and control quantity change range corresponding to each candidate process control sequence; Under the conditions that cooling output and heating output are mutually exclusive, and the temperature change rate and circulation flow rate meet preset constraints, candidate process control sequences that meet preset optimization conditions in terms of comprehensive evaluation value are selected.

5. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 1, characterized in that, In S5, the residual heat release process includes: Reduce the output corresponding to the current cooling and heating operating state and then stop the cooling and heating output; Adjust the circulation flow rate and flow path distribution based on the marked residual heat state and the temperature margin between the process object temperature and the corresponding temperature boundary. Predict the temperature change caused by residual heat and activate the opposite hot and cold working state when the start-up conditions are met.

6. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 5, characterized in that, The activation conditions include: The absolute value of the signed residual heat state is not greater than the residual heat threshold; Predicting temperature changes will not cause the temperature of the process object to exceed the corresponding temperature boundary. Without initiating the opposite hot / cold operating state, the temperature of the process object cannot enter the dynamic temperature tolerance range within a preset time.

7. The energy consumption optimization process control method for an industrial cooling and heating temperature control unit according to claim 1, characterized in that, In S5, the update includes: Obtain the temperature prediction deviation between the actual temperature of the process object and the predicted temperature of the process object; Based on temperature prediction deviations and process control data, update the equivalent heat capacity, heat transfer coefficient, transport hysteresis time, and residual heat retention coefficient. If the temperature prediction deviation exceeds the preset threshold, the rolling optimization is re-executed.

8. An energy consumption optimization process control system for an industrial cooling and heating temperature control unit, based on the control method described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to acquire process control data and process temperature programs; The generation module is used to determine the dynamic temperature tolerance range; The observation module is used to determine the state of the signed residual heat. The prediction module is used to predict the future temperature of the process object and the energy consumption of the process. The optimization module is used to obtain the target process control sequence under the constraints of the dynamic temperature tolerance domain; The control module is used to determine the residual heat release process and the start-up conditions for the opposite hot and cold operating states; The output module is used to output process control commands corresponding to refrigeration, heating, circulation flow, and flow path allocation; The update module is used to update the temperature dynamics model and residual heat state.

9. The energy consumption optimization process control system for an industrial cooling and heating temperature control unit according to claim 8, characterized in that, The acquisition module includes a process object temperature sensor, a liquid supply temperature sensor, a liquid return temperature sensor, a flow detector, and an energy input detector; the output module includes a cooling output control interface, a heating output control interface, a circulating flow control interface, and a flow path distribution control interface; each module is located in the process controller.

10. The energy consumption optimization process control system for an industrial cooling and heating temperature control unit according to claim 9, characterized in that, It also includes an internal buffer flow path and a flow path switching mechanism connected in parallel with the heat exchange flow path of the process object; the control module is used to control the flow path switching mechanism according to the marked residual heat state and temperature margin, and prohibits the output of cooling or heating process control commands with opposite polarity before the start-up conditions are met.

Images