Variable air volume painting workshop state machine and variable parameter temperature and humidity decoupling method

By constructing a parameter mapping model and online load identification in the air conditioning system of the painting workshop, and combining a multi-dimensional state machine and a decoupling compensation algorithm, the problem of temperature and humidity decoupling failure in the air conditioning system of the painting workshop under the conditions of variable air volume and spraying wet load changes was solved, and efficient and stable control of the painting workshop environment was achieved.

CN123067640APending Publication Date: 2026-11-10COSCO SHIPPING (QIDONG) OFFSHORE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611569887.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-12
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing air conditioning systems in painting workshops suffer from problems such as temperature and humidity decoupling failure, control response lag, and coupling oscillation when faced with variable air volume and large changes in spraying wet load. In particular, the dynamic abrupt changes in spraying wet load and the lag in sensor load status perception lead to poor control performance.

Method used

A variable air volume painting workshop state machine and a variable parameter temperature and humidity decoupling method are adopted. By offline calibration of the four-channel air thermodynamic transfer parameters, a parameter mapping model is constructed. Online load identification is performed by combining the spray gun trigger signal and exhaust volatile concentration data. Using a multi-dimensional finite state machine and decoupling compensation algorithm, the control parameters are dynamically reconstructed to suppress temperature and humidity coupling oscillations.

Benefits of technology

It achieves precise control of temperature and humidity in the painting workshop, reduces load sensing response lag, suppresses temperature and humidity coupling oscillations during working condition switching and production changes, and improves the stability of the painting workshop environment and the quality of spraying.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123067640A_ABST
    Figure CN123067640A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of coating workshop air conditioning control, and specifically discloses a variable air volume coating workshop state machine and a variable parameter temperature and humidity decoupling method, which is applied to a coating workshop control system configured with a variable air volume air conditioning unit, wherein the control system is connected with an execution valve group and comprises the following steps: obtaining multiple groups of offline air volume and spraying wet load working condition points, identifying air thermodynamic four-channel transmission parameters under each working condition point through a step response method, constructing a parameter mapping model with relative air volume and relative total load as independent variables, and generating a control strategy library. In the operation of the coating workshop, according to the offline calibrated thermodynamic parameter mapping model, the real-time total spraying wet load is identified by combining the online collection of spraying gun actions and the concentration of exhaust volatile substances, so that the hysteresis of load sensing response is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of air conditioning control technology in painting workshops, and particularly to a state machine for a variable air volume painting workshop and a variable parameter temperature and humidity decoupling method. Background Technology

[0002] When variable air volume (VAV) air conditioning systems are applied in painting workshops, they adapt to load changes during the production process by adjusting the air supply volume. Existing temperature and humidity control methods in painting workshops often employ single-loop proportional-integral-derivative feedback control, or set fixed feedforward decoupling parameters for temperature and humidity regulation under specific operating conditions.

[0003] In the coating process, the surface cooling and steam humidification processes of the air conditioning system are subject to cross-coupling of temperature and humidity. Existing fixed-parameter decoupling methods are ill-suited to the changes in thermodynamic properties caused by simultaneous and significant variations in airflow and multi-station spraying loads. Furthermore, the spraying wet load exhibits dynamic and abrupt changes, and existing systems rely on passive feedback from exhaust sensors, making it difficult to obtain the true load status in real time. This leads to problems such as temperature and humidity decoupling failure, control response lag, and coupling oscillations under variable airflow and load conditions. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in related technologies. Therefore, the purpose of this invention is to propose a state machine for a variable air volume painting workshop and a variable parameter temperature and humidity decoupling method to suppress the temperature and humidity coupling oscillation phenomenon that easily occurs during operating condition switching and changes in production conditions.

[0005] To achieve the above objectives, a first aspect of the present invention proposes a state machine for a variable air volume (VAV) painting workshop and a variable parameter temperature and humidity decoupling method, applied to a control system for a painting workshop equipped with a VAV air conditioning unit. The control system is connected to an actuator valve group and includes:

[0006] Multiple sets of offline air volume and spraying wet load operating points are obtained. The air thermodynamic four-channel transfer parameters under each operating point are identified by the step response method. A parameter mapping model with relative air volume and relative total load as independent variables is constructed to generate a control strategy library.

[0007] Collect the spray gun trigger signals and volatile concentration data of each spraying station and the exhaust manifold. Use a soft measurement algorithm to identify and calculate the preliminary wet load of each spraying station and the real-time total spraying wet load of the workshop online, and calculate the real-time relative total load.

[0008] The scheduling command, real-time air supply volume, duct static pressure and the real-time total wet spraying load of the workshop are used as four-dimensional trigger conditions to input into the multi-dimensional finite state machine. When it is determined that a state cross-interval switch occurs, the corresponding benchmark parameter set is output. When it is fluctuating in the same interval, stepless parameter interpolation is performed.

[0009] Using real-time relative air volume and real-time relative total load as dual scheduling variables, the parameter mapping model is invoked to reconstruct the temperature and humidity cross-feedforward decoupling coefficient and the operating parameters of the main loop controller online. The main loop controller includes a temperature loop and a humidity loop.

[0010] By combining the reconstructed temperature and humidity cross-feedforward decoupling coefficient with the operating parameters, a decoupling compensation calculation is performed to generate an opening control signal to control the cold water valve and humidification valve in the actuator valve group.

[0011] To achieve the above objectives, a second aspect of the present invention provides a state machine and a variable parameter temperature and humidity decoupling system for a coating workshop, applied to a control system for a coating workshop equipped with a variable air volume air conditioning unit. The control system is connected to an actuator valve group and includes:

[0012] The offline calibration and modeling module is used to obtain multiple sets of offline air volume and spraying wet load operating points, identify the air thermodynamic four-channel transfer parameters under each operating point through the step response method, construct a parameter mapping model with relative air volume and relative total load as independent variables, and generate a control strategy library.

[0013] The online load identification module is used to collect the spray gun trigger signals and volatile concentration data of each spraying station and the exhaust manifold. It uses a soft measurement algorithm to identify and calculate the preliminary wet load of each spraying station and the real-time total spraying wet load of the workshop, and calculates the real-time relative total load.

[0014] The state machine scheduling module is used to input the scheduling command, real-time air supply volume, duct static pressure and the real-time total spraying wet load of the workshop as four-dimensional trigger conditions into the multi-dimensional finite state machine. When it is determined that a state cross-interval switch occurs, the corresponding reference parameter set is output. When it is fluctuating within the same interval, stepless parameter interpolation is performed.

[0015] The parameter online reconstruction module is used to call the parameter mapping model with real-time relative air volume and real-time relative total load as dual scheduling variables to reconstruct the temperature and humidity cross-feedforward decoupling coefficient and the operating parameters of the main loop controller online. The main loop controller includes a temperature loop and a humidity loop.

[0016] The decoupling compensation control module is used to perform decoupling compensation calculations by combining the reconstructed temperature and humidity cross-feedforward decoupling coefficient with the operating parameters to generate an opening control signal to control the cold water valve and humidification valve in the actuator valve group.

[0017] To achieve the above objectives, a third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory. When the computer program is executed by the processor, it implements the above-described variable air volume painting workshop state machine and variable parameter temperature and humidity decoupling method.

[0018] The technical solution of this application, in the operation of the painting workshop, identifies the real-time total wet spraying load based on an offline calibrated thermodynamic parameter mapping model, combined with online acquisition of spray gun actions and exhaust volatile concentration, thereby reducing the lag in load perception response. The control system, combined with multi-dimensional trigger conditions, schedules the output reference parameters of the finite state machine and reconstructs the feedforward decoupling coefficient and operating parameters online using real-time air volume and load as variables. This decoupling compensation process enables the control actuator to follow the actual air volume fluctuations and workstation operation rhythm, dynamically offsetting the cross-interference effects of surface cooling and steam humidification, and suppressing the temperature and humidity coupling oscillation phenomenon that is prone to occur when switching operating conditions and changing production status. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the variable air volume painting workshop state machine and variable parameter temperature and humidity decoupling method provided by the present invention.

[0020] Figure 2 This is a schematic diagram illustrating the implementation of the variable air volume painting workshop state machine and variable parameter temperature and humidity decoupling system provided by the present invention.

[0021] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0022] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0023] The following description, with reference to the accompanying drawings, describes the variable air volume painting workshop state machine, variable parameter temperature and humidity decoupling method, system, and electronic equipment according to embodiments of the present invention.

[0024] Example 1:

[0025] like Figure 1 As shown, this embodiment provides a state machine for a variable air volume (VAV) painting workshop and a variable parameter temperature and humidity decoupling method. This method deeply integrates the complex aerodynamic and thermodynamic characteristics of modern painting workshops and is applied to the control system of a painting workshop equipped with a VAV air conditioning unit.

[0026] Specifically, the control system is connected to an actuator valve assembly via fieldbus or hardwiring. The actuator valve assembly is the physical terminal for regulating air temperature and humidity, primarily comprising a chilled water valve for introducing chilled water to lower air temperature and condensate, and a humidifying valve for injecting high-temperature steam to increase absolute humidity. The air conditioning system in the painting workshop achieves variable air volume operation by adjusting the inverter outputs of the main supply fan and exhaust fan, thereby significantly reducing energy consumption while meeting the requirements of the painting process.

[0027] However, variable air volume operation can cause drastic changes in the thermodynamic transfer inertia of the system. The method provided in this embodiment aims to solve the problem of temperature and humidity coupling oscillation caused by this dynamic change.

[0028] Specifically, the method of the present invention includes the following steps:

[0029] Step 1: The method in this embodiment first requires performing an offline benchmark calibration phase. During this phase, the system acquires multiple sets of offline airflow and spraying wet load operating points.

[0030] In the implementation of this project, the air supply volume in the painting workshop is not arbitrarily and continuously selected, but rather divided into multiple discrete typical operating ranges based on process requirements. Similarly, the spraying wet load also exhibits a stepped variation depending on the number of spraying stations operating simultaneously. Therefore, the system is designed with multiple sets of cross-operating point conditions for air volume and wet load to cover the entire normal operating envelope of the painting workshop.

[0031] After determining the aforementioned operating points, the system identifies the four-channel air thermodynamic transfer parameters at each operating point using the step response method. The step response method involves artificially applying a step opening change signal to a specific actuator (such as a chilled water valve or a humidifier valve alone) when the air conditioning system is in steady-state equilibrium, and simultaneously recording the dynamic response curves of the supply air temperature and humidity over time. By performing tangent fitting or area integration on this dynamic response curve, the transfer function characteristics of the system under that specific airflow and load are identified.

[0032] Since the air conditioning system in the painting workshop is a typical dual-input dual-output system, the four-channel air thermodynamic transfer parameters specifically include the static gain of the temperature main channel, the static gain of the humidity main channel, the static gain of the temperature-induced humidity coupling channel, and the static gain of the humidity-induced temperature coupling channel.

[0033] Here, the static gain of the temperature main channel is defined as the steady-state temperature change caused by a unit change in the opening degree of the chilled water valve; the static gain of the humidity main channel is defined as the steady-state absolute humidity change caused by a unit change in the opening degree of the humidifier valve; the static gain of the temperature-humidity coupling channel is defined as the cross-effect gain of the decrease in absolute air humidity caused by the cooling process of the chilled water valve due to the surface temperature of the surface cooler being lower than the dew point and condensation of water; the static gain of the humidity-temperature coupling channel is defined as the cross-effect gain of the increase in dry-bulb air temperature caused by the release of sensible heat from the steam when the humidifier valve injects high-temperature steam to increase humidity.

[0034] After identifying all offline operating points, the system constructs a parameter mapping model with relative airflow and relative total load as independent variables, generating a control strategy library. The relative airflow refers to the ratio of the current actual airflow to the system's designed rated airflow; the relative total load refers to the ratio of the current actual total wet load to the system's designed rated total wet load. Using the corresponding parameters under rated airflow and rated wet load as benchmark values, the system employs a bivariate quadratic polynomial structure to fit the four-channel air thermodynamic transfer parameters.

[0035] In three-dimensional space, this fitting process is equivalent to approximating the scattered data obtained offline with a smooth bivariate quadratic surface, thereby calculating accurate thermodynamic parameters even at intermediate operating points that are not directly tested. The general expression of the generated parameter mapping model is shown in formula (1):

[0036] (1)

[0037] In formula (1), Defined as when the relative air volume is And the relative total load is The calculated value of the target thermodynamic parameter is output under real-time conditions, which can represent any one of the static gains of the four channels mentioned above. During the real-time control execution cycle, the system will use this calculated value. The value is assigned to the underlying physical execution register as the actual output gain at the current moment; Defined as the reference value of the target thermodynamic parameter under the corresponding rated calibration conditions; Defined as relative air volume, it is a dimensionless percentage value; Defined as a relative total load, it is also a dimensionless percentage value; , , , , , All are dimensionless mapping coefficients generated offline by fitting step response data using the least squares method. These represent the model constants. These six dimensionless mapping coefficients fully characterize the nonlinear distortion effect of airflow and load variations on the thermodynamic transfer properties of the painting workshop. This control strategy library is stored non-volatilely in the underlying memory of the control system.

[0038] Step 2: After the painting workshop is officially put into online production, the system needs to obtain the workshop's load status in real time.

[0039] Physical humidity sensors are prone to condensation, blockage, and measurement drift in exhaust ducts filled with paint mist and solvent volatiles, leading to severe distortion in load perception. Therefore, this embodiment collects the spray gun trigger signals from each spraying station and the volatile concentration data from the main exhaust duct. A soft measurement algorithm is then used to identify and calculate the preliminary wet load of each spraying station and the real-time total wet load of the workshop online, and to calculate the real-time relative total load.

[0040] The soft measurement algorithm described is a technique that uses easily measurable and rapidly responding auxiliary variables, combined with a mathematical model, to estimate target variables that are difficult to measure directly. In this step, the initial wet load of each spraying station is first calculated by combining the pre-calibrated single-gun rated wet generation rate of each spraying station with the spray gun trigger signal. The single-gun rated wet generation rate refers to the evaporation mass of solvent and water contained in the paint sprayed by the spray gun per unit time; the spray gun trigger signal is a digital switch signal. When the spray gun trigger signal is high, the high-level time is multiplied by the single-gun rated wet generation rate to obtain the initial wet load of that specific spraying station. Subsequently, the initial wet loads of all parallel-operating spraying stations are algebraically summed to generate the total initial wet load.

[0041] However, the initial total wet load calculated solely based on the spray gun trigger signal is a feedforward estimate and does not account for uncertainties caused by fluctuations in paint transfer rate and human differences in spraying techniques. Therefore, feedback correction is required. Volatile organic compounds (VOCs) in the paint evaporate simultaneously with moisture during the spraying process. Therefore, the system uses the VOC concentration data from the exhaust duct as the observation benchmark and employs a recursive least squares algorithm with a forgetting factor to calculate the overall load identification correction coefficient.

[0042] The core of the recursive least squares algorithm lies in calculating the difference between the actual volatile concentration collected at the current moment and the theoretical volatile concentration predicted based on the total preliminary wet load, thus obtaining the concentration prediction error. Subsequently, this concentration prediction error, combined with the internal covariance matrix, is used to iteratively update the overall load identification correction coefficient from the previous moment. The introduction of a forgetting factor aims to gradually reduce the weight of historical data on the current identification result, enabling the algorithm to possess excellent time-varying tracking capabilities and quickly adapt to the dynamic changes in the painting workshop's process cycle.

[0043] After iteratively obtaining the overall load identification correction coefficient, the system multiplies the total preliminary wet load by the overall load identification correction coefficient to eliminate the static bias of the feedforward estimation and obtain the high-precision and corrected real-time total spraying wet load of the workshop. Further, the system divides the real-time total spraying wet load of the workshop by the rated total wet load baseline value determined at the beginning of the workshop design to generate the real-time relative total load for subsequent control and scheduling.

[0044] Step 3: In order to ensure the uniformity of airflow organization inside the painting workshop and avoid the diffusion of paint mist in high-load workstations and the waste of fresh air in low-load workstations, the method also includes the step of allocating air supply branches as needed.

[0045] The system first determines whether the total preliminary wet load is greater than a preset zero-point dead zone threshold. If it is greater than the zero-point dead zone threshold, the system calculates the proportion of the preliminary wet load of each spraying station in the total preliminary wet load; if it is less than or equal to the zero-point dead zone threshold, the system forcibly sets the proportion to an equal default value to prevent division-by-zero algorithm failures and maintain the micro-positive pressure balance of the workshop foundation. The physical meaning of this proportion is to characterize the spatial distribution weight of pollution sources inside the workshop at the current moment.

[0046] Next, according to the aforementioned proportional coefficient, the real-time total wet load of the workshop identified and corrected by soft measurement is decomposed to obtain the real-time wet load of each of the spraying stations. Subsequently, the system uses the real-time wet load of each station as the allocation weight to distribute the total air volume provided by the main air duct to each air supply branch according to the weight ratio, and calculates the target required air volume for each air supply branch. This on-demand allocation mechanism ensures, from an aerodynamic perspective, that the air volume in a local area matches its instantaneous pollutant generation, significantly improving paint mist removal efficiency.

[0047] Step 4: Production in the painting workshop is not a static steady state, but rather involves multiple typical process operation stages. The system inputs the scheduling instructions, real-time air supply volume, duct static pressure, and the workshop's real-time total wet spraying load as four-dimensional trigger conditions into a multi-dimensional finite state machine. When a state transition across intervals is detected, the system outputs the corresponding set of reference parameters; when the state fluctuates within the same interval, it performs stepless parameter interpolation.

[0048] The multidimensional finite state machine is a logical judgment model that includes five discrete process states: hibernation, transition, standby, spraying, and self-cleaning. Specifically:

[0049] The "dormant" state corresponds to non-production shifts, where the system maintains only a very low airflow to ensure a basic positive pressure; the "transition" state corresponds to the preparation stage when the workshop is heating up or cooling down; the "standby" state corresponds to production breaks, where the airflow is maintained at a low level; the "spraying" state corresponds to the workpiece loading stage, where the airflow and load reach the normal range; and the "self-cleaning" state corresponds to the strong exhaust cleaning stage after spraying, where the airflow is at its maximum but the load drops sharply.

[0050] Because the coating process has stringent environmental requirements, any state transition can cause a reconstruction of the airflow organization. Therefore, triggering the state transition across intervals requires the strict fulfillment of the four-dimensional triggering conditions:

[0051] First, the scheduling instructions received from the host manufacturing execution system must match the target operating conditions.

[0052] Second, the real-time airflow rate fed back by the frequency converter must remain within the airflow range of the target operating condition for a first preset time. The first preset time is set to filter out sensor reading pulses caused by short-term aerodynamic airflow turbulence.

[0053] Third, the static pressure of the duct measured by the differential pressure sensor installed on the main pipeline must fluctuate within the set static pressure error range. This indicates that the fan impeller has broken away from the dynamic process of acceleration or deceleration, and the pneumatic pipeline network has re-established static pressure balance.

[0054] Fourth, the total real-time wet spraying load in the workshop, as identified by soft measurement, remains within the load range of the target operating condition for a second preset time. The second preset time is usually longer than the first preset time to confirm that there is indeed continuous and stable spraying operation at the workstation, rather than a short-term test spraying operation.

[0055] Furthermore, in order to enhance the robustness of the finite state machine and prevent frequent logic jumps in the system at the boundary of the operating conditions, the airflow exit threshold and load exit threshold set for the state cross-interval switching are respectively lower than the corresponding airflow entry threshold and load entry threshold, forming an anti-shake hysteresis interval.

[0056] For example, if the airflow threshold for entering the "spraying" state is determined to be 80% of the rated airflow, then the airflow exit threshold for exiting the "spraying" state and falling to the "standby" state may be set to 75%. This 5% difference constitutes the hysteresis band, effectively suppressing control oscillations caused by system noise.

[0057] It is also important to note that the air conditioning units in large painting workshops contain a huge water volume within their coils, resulting in a significant lag in their thermodynamic response. Therefore, the operation of the multidimensional finite state machine also includes feedforward operations based on scheduling.

[0058] The system collects workpiece queue and scheduling plan data from the production line via industrial Ethernet and predicts the direction of work condition switching using a sliding window pattern matching algorithm. Specifically, the system arranges the workpiece arrival sequence for the next five minutes within a time sliding window and, by comparing it with a preset workpiece type and paint consumption model, anticipates that the system is about to transition from "standby" to a high-load "spraying" state.

[0059] When the output condition switching prediction signal is received, before all four-dimensional trigger conditions are met, the system proportionally feeds forward the current set of reference parameters to the parameter setting value corresponding to the target condition. This action is equivalent to fine-tuning the opening of the chilled water valve and humidification valve before the actual physical command arrives, using the equipment's hysteresis time to offset the delay in physical response and achieve a smooth condition transition.

[0060] Step 5: After the multidimensional finite state machine determines the current discrete state, the system performs stepless interpolation for minor parameter fluctuations within the state interval. This method uses real-time relative airflow and the real-time relative total load as dual scheduling variables, calls the parameter mapping model, and reconstructs online the temperature and humidity cross-feedforward decoupling coefficients and the operating parameters of the main loop controller. The main loop controller includes a temperature loop and a humidity loop.

[0061] The core of the above reconstruction process lies in decoupling operations. As mentioned earlier, air conditioning units exhibit strong physical coupling when regulating temperature and humidity. The real-time relative airflow is used as... And the real-time relative total load is taken as Substituting into the parameter mapping model of formula (1), the system can calculate the extremely accurate static gain of the temperature main channel, the static gain of the humidity main channel, the static gain of the temperature-induced humidity coupling channel, and the static gain of the humidity-induced temperature coupling channel under the current transient conditions.

[0062] Subsequently, based on the principle of superposition of linear systems, the temperature and humidity cross-feedforward decoupling coefficients are calculated.

[0063] First, the ratio of the static gain of the temperature-induced humidity coupling channel to the static gain of the humidity main channel is negatively adjusted to generate a temperature-induced humidity feedforward decoupling coefficient used to offset the interference of surface cooling, condensation, and dehumidification. Its calculation is shown in formula (2):

[0064] (2)

[0065] In formula (2), Defined as the generated temperature-humidification feedforward decoupling coefficient, its physical meaning is the relative proportion of the humidification valve's synchronous compensation action when the cold water valve opens, so that the cooling process does not affect the absolute humidity of the air. Defined as the static gain of the wet-induced wet coupling channel calculated from the parameter mapping model; Defined as the static gain of the main humidity channel calculated from the parameter mapping model; the negative sign indicates that the direction of the compensation action must be opposite to the direction of the effect of the interference.

[0066] Secondly, the ratio of the static gain of the humidification coupling channel to the static gain of the temperature main channel is negatively calculated to generate a humidification feedforward decoupling coefficient used to counteract the temperature rise interference from steam humidification. Its calculation is shown in formula (3):

[0067] (3)

[0068] In formula (3), Defined as the generated wet-induced temperature feedforward decoupling coefficient, its physical meaning is that when the humidifying valve operates to spray steam, the cold water valve needs to be opened in a synchronously compensated proportion to eliminate the temperature rise disturbance caused by the sensible heat of the steam. Defined as the static gain of the wet-induced temperature coupling channel calculated from the parameter mapping model; Defined as the static gain of the main temperature channel calculated from the parameter mapping model.

[0069] Step Six: After obtaining the accurate feedforward decoupling coefficients, the system combines the reconstructed temperature and humidity cross-feedforward decoupling coefficients with the operating parameters to perform decoupling compensation calculations and generate opening control signals to control the cold water valve and humidification valve in the actuator valve group.

[0070] The system first performs standard positional or incremental proportional-integral-derivative (PID) calculations in the temperature and humidity loops, respectively, based on the reconstructed operating parameters (including proportional gain, integral time, and derivative time). In the temperature loop, an initial temperature control variable is calculated based on the deviation between the current measured supply air temperature and the setpoint; in the humidity loop, an initial humidity control variable is calculated based on the deviation between the current measured absolute humidity and the setpoint. These two control variables are independent, non-decoupled calculation results.

[0071] Subsequently, a feedforward cross-addition operation is performed. The product of the initial temperature control quantity and the temperature-humidification feedforward decoupling coefficient is used as the first compensation quantity and added to the initial humidity control quantity to generate the final opening control signal for controlling the humidification valve. Its calculation is shown in formula (4):

[0072] (4)

[0073] In formula (4), Defined as the opening control signal output to the humidifier valve's electric positioner; Defined as the initial humidity control value calculated by the humidity loop; Defined as the initial temperature control quantity calculated by the temperature loop.

[0074] Simultaneously, the product of the initial humidity control quantity and the humidity-induced temperature feedforward decoupling coefficient is added to the initial temperature control quantity as a second compensation quantity to generate the final opening control signal for the chilled water valve. Its calculation is shown in formula (5):

[0075] (5)

[0076] In formula (5), Defined as the opening control signal of the cold water valve output to the cold water valve electric positioner. Through the above calculations, the system achieves mathematical decoupling of two physically strongly coupled actuators at the algorithm level, ensuring that any unidirectional adjustment request for temperature or humidity will not cause drastic fluctuations in the other variable.

[0077] Step 7: After obtaining the target air volume requirement for each air supply branch in Step 3 above, the target air volume requirement may deviate from the actual result due to the difference in friction coefficient of the physical duct system.

[0078] Therefore, the system collects the actual air volume measured at the air volume regulating valve at the end of each air supply branch, and calculates the branch air supply balance index by combining the target required air volume with the set air volume error lower limit.

[0079] The calculation method for the branch air supply balance index is as follows: subtract the actual air volume of each air supply branch from the corresponding target air demand volume to obtain the air volume deviation of a single branch, calculate the square value of the air volume deviation of the single branch, and take the arithmetic mean of the square values ​​of all air supply branches to obtain the branch air supply balance index that characterizes the overall airflow distribution balance of the workshop.

[0080] The control system is equipped with a first air supply deviation threshold and a second air supply deviation threshold, and the first air supply deviation threshold is strictly defined as being greater than the second air supply deviation threshold.

[0081] When the branch air supply balance index is determined to be greater than the first air supply deviation threshold, it indicates that the overall airflow distribution in the workshop is severely unbalanced, requiring the activation of a forced balancing strategy. For air supply branches where the actual air volume deviates from the target air volume within a preset upper deviation range, the system increases the pulsation frequency of the corresponding damper stepper motor to accelerate the damper's closing speed; and reduces the static pressure compensation angle of the associated fan to decrease the local dynamic pressure conversion in that branch. Conversely, for air supply branches where the actual air volume deviates from the target air volume within a preset lower deviation range (i.e., the actual air volume is severely insufficient), the system reduces the damper pulsation frequency to prevent over-adjustment and expands the static pressure compensation angle to actively increase the air supply static pressure source of that branch. Through this directional adjustment based on regional differences, air volume deviations in each branch can be quickly eliminated.

[0082] Step 8: It is also important to note that the heat load of the painting workshop is not only affected by normal spraying operations, but also by strong external disturbances such as the opening and closing of workshop doors and sudden changes in outdoor seasonal weather conditions. Therefore, the method also includes a load adaptive disturbance rejection control step.

[0083] The system first calculates a dynamic warning threshold based on the real-time relative total load. This calculation logic follows a core engineering principle: the dynamic warning threshold decreases proportionally as the real-time relative total load increases. This is because in a painting workshop under high relative load conditions (such as full-load spraying), the workpiece has extremely low tolerance for temperature and humidity precision, therefore the warning threshold must be tightened; while under low relative load conditions (such as standby preparation), the warning threshold can be appropriately relaxed to reduce frequent wear on the valve's mechanical structure.

[0084] During operation, when the absolute value of the temperature and humidity control deviation is less than the dynamic early warning threshold, the system determines that it is currently in a normal and controllable range, maintains the control parameters of the decoupling compensation operation unchanged, and executes step six in a routine manner.

[0085] When the absolute value of the temperature and humidity control deviation is greater than or equal to the dynamic warning threshold, the system determines that it has been subjected to a strong external disturbance, and the conventional feedforward decoupling strength is insufficient to suppress the fluctuation. At this time, the system immediately triggers the enhanced anti-disturbance mode: the temperature and humidity cross-feedforward decoupling coefficient calculated in step five above is multiplied by a first proportional coefficient greater than 1 for forced amplification, thereby increasing the adjustment strength of cross-compensation; at the same time, the proportional band in the operating parameters is multiplied by a second proportional coefficient less than 1 for forced narrowing (reducing the proportional band is equivalent to increasing the controller's proportional gain), improving the controller's response speed and adjustment strength to the deviation, thereby controlling the deviation within the safe set threshold in a shorter time.

[0086] If the disturbance is too extreme, causing the calculated valve opening control signal to reach 100% full opening or 0% full closing, i.e., the valve assembly is in a saturated opening limit state, without intervention, it will lead to deep integral saturation and control system logic chaos. At this time, the system immediately intervenes with an abnormal fallback control program: forcibly locking the state switching process of the multidimensional finite state machine to prevent erroneous logic jumps caused by data distortion; at the same time, strictly limiting the output change rate of the opening control signal to within a set safe change rate threshold (e.g., no more than 2% opening change per second). This smoothing and limiting method prevents water hammer or steam jet impact caused by valve overshoot, ensuring the safe operation of the underlying physical equipment.

[0087] Based on existing technologies, traditional air conditioning control methods in painting workshops often employ single-loop PID control with fixed parameters or only tune a set of static decoupling parameters at specific rated points. When faced with variable airflow regulation and significant fluctuations in the wet load of spraying stations, which are common in painting processes, these methods inevitably fail due to a lack of real-time sensing of air thermodynamic characteristics and dynamic parameter mapping. This results in severe cross-coupling oscillations in the temperature and humidity control curves. Furthermore, traditional systems heavily rely on passive feedback from humidity sensors inside the exhaust ducts, leading to a significant lag in load perception response.

[0088] The comprehensive technical solution described in this embodiment first establishes an offline calibrated binary two-dimensional parameter mapping model, enabling the control system to calculate extremely accurate thermodynamic transfer gain across the entire operating envelope. Subsequently, by employing a soft-sensor identification algorithm based on exhaust volatile concentration and spray gun trigger signal, the hysteresis of physical humidity sensors is eliminated, achieving pre-sensing of the total spraying wet load.

[0089] Furthermore, a multidimensional finite state machine is used to systematically schedule the complex process stages in the workshop. Operating parameters and feedforward decoupling coefficients are reconstructed online when airflow and load change, and rigorous decoupling compensation calculations are performed on the final valve position output signal. Finally, branch balancing and a dynamic adaptive disturbance rejection fallback mechanism are employed.

[0090] The entire technical solution is interdependent and interconnected. From data acquisition and model prediction to logic scheduling and actuator intervention, it forms a highly closed-loop intelligent control link. This effectively counteracts the cross-interference effects of physical actuators and suppresses the temperature and humidity coupling oscillation phenomenon that is prone to occur when switching working conditions and changing production conditions. This effectively improves the stability of the high-precision working environment in the painting workshop and the quality of product spraying.

[0091] Example 2:

[0092] Throughout the entire lifecycle of a painting workshop, there exists a type of "slow time-varying disturbance" that cannot be ignored. Specifically, the exhaust system of a painting workshop is equipped with multi-stage paint mist filters or water-washing collection devices to intercept overspray paint mist that does not adhere to the surface of the workpiece during the spraying operation.

[0093] As production continues, a large amount of solid paint mist particles and resin components will continuously adhere to and gradually clog the pores of the paint mist filter. This physical clogging and dust accumulation process causes the aerodynamic resistance of the exhaust duct network to slowly but continuously increase over time. The increase in duct network resistance prolongs the pure time lag of the system's aerodynamic response, and alters the retention and mixing characteristics of air in the surface cooler and steam humidification section. This leads to deviations in the parameter mapping model obtained from the offline calibration of the clean system in Example 1, causing a gradual loss of the phase margin of the closed-loop control system.

[0094] To address the control law degradation problem caused by the long-term deterioration of physical equipment, this embodiment further introduces a dynamic hysteresis compensation mechanism based on Embodiment 1.

[0095] Specifically, the method also includes a hysteresis compensation step for paint mist filter resistance degradation caused by slow time-varying aerodynamic drag disturbances. The primary step in this process is the accurate perception of the physical attenuation state of the exhaust system and the mathematical extraction of its characteristic parameters.

[0096] At the engineering implementation level, the control system acquires real-time differential pressure data of the paint mist filter at the exhaust system via an industrial fieldbus. Typically, a high-precision differential pressure transmitter is installed between the inlet and outlet sides of the paint mist filter on the exhaust fan unit. Considering that the exhaust fan generates high-frequency airflow pulsations and eddy current noise during operation due to impeller rotation, the directly acquired differential pressure signal often exhibits significant spikes and high-frequency jitter.

[0097] Therefore, the control system pre-embeds a moving average filtering algorithm in the underlying data acquisition module to smooth the original differential pressure sampling points within a set time window, so as to filter out high-frequency aerodynamic noise and extract a steady-state differential pressure signal that can truly reflect the degree of physical blockage of the filter screen pores.

[0098] After obtaining the filtered real-time steady-state differential pressure, the use of absolute differential pressure values ​​for control law correction is not universally applicable due to the varying paint mist filter materials (such as dry cardboard boxes, fiberglass wool, or wet venturi tubes) and design rated differential pressures in painting workshops of different sizes and production lines. Furthermore, it is difficult to directly interface with existing standardized control strategy libraries. Therefore, the system divides the paint mist filter differential pressure data by a preset initial clean differential pressure in its software logic to calculate a dimensionless resistance degradation index.

[0099] The initial clean pressure difference refers to the baseline pressure difference measured under steady-state operation conditions with rated air supply and exhaust volume after the initial equipment commissioning in the painting workshop or after a brand new paint mist filter has been replaced during the maintenance cycle. This baseline value is entered and stored in non-volatile memory during the initial configuration phase of the control system. The specific mathematical expression for calculating the dimensionless resistance degradation index is shown in formula (6):

[0100] (6)

[0101] In formula (6), Defined as the dimensionless drag degradation index generated by calculation; Defined as the real-time pressure difference data of the paint mist filter after digital filtering; Defined as the initial clean pressure differential preset by the system.

[0102] Through this division operation mechanism, the system successfully abstracts the pressure difference physical quantity with specific engineering dimensions into a dimensionless state variable that purely characterizes the proportion of aerodynamic drag degradation. The construction of this dimensionless exponent allows the subsequent compensation logic to be freed from the constraints of specific hardware selection, and has wide reusability across multiple coating production lines using different filtration equipment.

[0103] Optionally, after calculating the drag degradation index in real time, the control system does not immediately trigger modifications to the control parameters for any minor numerical fluctuations. Since the paint shop's air supply and exhaust frequency converter system itself has a certain degree of disturbance rejection redundancy, the slight dust accumulation on the paint mist filter during the initial use phase has an impact on the overall aerodynamic inertia of the system within a tolerable control dead zone. To prevent the compensation system from becoming overly sensitive and causing unnecessary parameter fluctuations, a dead zone mechanism is introduced into the control system's logic judgment stage.

[0104] When the resistance degradation index exceeds a preset degradation initiation threshold, a dynamic hysteresis penalty coefficient is generated through positive mapping based on the difference between the resistance degradation index and the degradation initiation threshold. The degradation initiation threshold is a critical dimensionless constant pre-determined through fluid dynamics simulation or historical operating data statistics. It indicates that the degree of clogging of the paint mist filter has reached a critical threshold affecting system performance, and the hysteresis effect of system exhaust causes the phase margin of the temperature and humidity control loop to continuously decrease.

[0105] Once the drag degradation index is determined to have crossed the critical point, the system activates the dynamic compensation calculation engine to perform a forward mapping operation. Forward mapping refers to the phenomenon where, as the drag degradation index continues to rise, the generated dynamic hysteresis penalty coefficient also shows a corresponding increasing trend, representing the physical reality that the system response is becoming increasingly slower. The mapping calculation process for this dynamic hysteresis penalty coefficient is shown in formula (7):

[0106] (7)

[0107] In formula (7), Defined as the dynamic hysteresis penalty coefficient generated by the forward mapping; The physical meaning of this maximum-value function lies in constructing a unidirectional clamping barrier at the algorithmic level to ensure stability within the drag degradation exponential. The degradation initiation threshold has not been exceeded. At that time, the penalty term is forcibly cleared to zero, making By keeping the baseline value of 1 constant, the risk of erroneous reverse compensation caused by small negative fluctuations in the signal is eliminated from the mathematical source. Defined as a pre-calibrated degenerate mapping slope constant; Defined as the resistance degradation index calculated by formula (6); Defined as the system's preset degradation startup threshold.

[0108] The mapping formula clarifies that when the drag degradation index does not exceed the degradation initiation threshold, the control logic does not perform the compensation calculation, and the penalty coefficient is equivalent to one. However, once the threshold is exceeded, the difference in magnitude will be amplified according to the slope constant, and after being superimposed on the base value of one, the final dynamic hysteresis penalty coefficient is formed. This coefficient directly quantifies the delay factor of the increase in system aerodynamic drag on the thermodynamic response time.

[0109] For example, after accurately calculating the dynamic hysteresis penalty coefficient, the control system enters the dynamic correction stage of the control law. The dynamic hysteresis penalty coefficient is input as a slow time-varying feedforward variable into the online reconstruction stage to compensate for the integral time constant and derivative time constant in the operating parameters.

[0110] In the technical solution of Embodiment 1, the online parameter reconstruction module has reconstructed the basic main loop controller operating parameters from the binary two-dimensional parameter mapping model based on the two rapidly changing variables, "real-time relative air volume" and "real-time relative total load". In this embodiment, at the end of the reconstruction stage, the control system connects a slow time-varying adjustment correction stage based on the "dynamic hysteresis penalty coefficient".

[0111] This processing architecture is highly targeted in control theory. The operating parameters of the main loop controller (typically including proportional gain, integral time constant, and derivative time constant) determine the dynamic characteristics of the controller in eliminating control deviations. When the resistance of the paint mist filter increases and the system's aerodynamic inertia increases significantly, the response of the temperature and humidity in the painting workshop to the actions of the cold water valve and humidification valve will become more sluggish. If the integral time constant and derivative time constant reconstructed based on the clean system in Example 1 are continued to be used, it will lead to serious control instability risks.

[0112] Specifically, based on the dynamic hysteresis penalty coefficient, the integral time constant of the main loop controller is increased proportionally, and the derivative time constant is decreased.

[0113] First, we analyze the adjustment logic of the integral time constant. The physical meaning of integral action is to eliminate the static steady-state error of the system, but excessive integral action (i.e., a small integral time constant) will cause the controller output to accumulate rapidly. Under conditions where the system response becomes sluggish, the actual measured values ​​of temperature or humidity cannot change quickly with the opening of the valve.

[0114] If the original integral time constant is kept unchanged, the integral term inside the controller will accumulate rapidly and in large quantities during the system response waiting period (i.e., integral saturation). When the controlled variable finally starts to respond and crosses the setpoint, due to the excessively large integral term accumulated in the early stage, the valve cannot retract in time, thus causing a large overshoot and continuous oscillation.

[0115] Therefore, the control system multiplies the baseline integral time constant of the reconstructed output in Example 1 by the dynamic hysteresis penalty coefficient (since the penalty coefficient is greater than 1), thereby increasing the actual effective integral time constant proportionally. This operation essentially weakens the intensity of the integral action, lengthens the integral response period, and allows the controller's action rhythm to match the already slowed-down physical system response rhythm, effectively suppressing overshoot caused by integral oversaturation.

[0116] Next, we analyze the adjustment logic of the differential time constant. The physical meaning of the differential action lies in making advance predictions and adjustments based on the changing trend (rate of change) of the controlled variable. However, under conditions of significantly increased aerodynamic drag and prolonged pure time lag in the system, the rate of change of temperature or humidity will become very weak and lagging.

[0117] Even more seriously, when the system response is extremely slow, the random electrical noise or minute airflow disturbances superimposed on the sensor measurement signal may numerically exceed the actual rate of change caused by valve action. If the original differential time constant is kept constant at this time, the controller is very likely to amplify these irrelevant noises, causing the valve assembly to jitter frequently and irregularly. This not only fails to provide proactive regulation but also accelerates the wear of the valve's mechanical structure and undermines the stability of the control loop.

[0118] Therefore, the control system divides the reference differential time constant of the reconstructed output in Example 1 by the dynamic hysteresis penalty coefficient, thereby reducing the actual effective differential time constant. This operation essentially actively converges the controller's differential advance prediction capability, ensuring the smoothness and stability of the slow system's output in noisy environments by reducing its sensitivity to small rates of change.

[0119] It is also important to note that the parameter compensation operation based on aerodynamic drag degradation described above is not isolated from the multidimensional finite state machine scheduling and temperature and humidity cross-feedforward decoupling mechanism in Example 1, but rather forms a deep logical fusion. The feedforward decoupling logic in Example 1 is mainly responsible for eliminating the physical cross-coupling between temperature control and humidity control in the spatial dimension, and responding to rapid changes in airflow and process load in the short time dimension; while the paint mist filter drag degradation hysteresis compensation step in this example focuses on tracking and offsetting the time-varying decay of the system transfer function caused by the physical aging of consumables over a longer time span. The two are fast and slow in the time scale, and complement each other at the control level.

[0120] Based on current technology, traditional air conditioning control systems in painting workshops typically treat aerodynamic characteristics as static, time-invariant systems. During the lifespan of the paint mist filter, as the filter gradually becomes clogged, the phase margin of the original control system is irreversibly lost. This forces on-site maintenance engineers to frequently intervene manually, either by manually retuning the PID control parameters to mitigate oscillations or by prematurely replacing the filter before it reaches its maximum dust holding capacity. This not only significantly increases labor costs for maintenance but also results in a serious waste of filter consumables.

[0121] The technical solution proposed in this embodiment extracts the hardware differential pressure signal reflecting exhaust resistance into a dimensionless resistance degradation index, and uses this as a slow time-varying feedforward variable closed loop to intervene in the online reconstruction of the main loop control parameters. This solution enables the variable air volume (VAV) coating air conditioning control system to have adaptive adjustment capabilities throughout its entire life cycle. Without manual intervention, the control system can dynamically perform smooth degradation compensation on the integral and derivative time constants based on the objective physical degradation process of the paint mist filter. This ensures that even under harsh operating conditions where the paint mist filter is nearing the end of its lifespan and the system aerodynamic resistance reaches its peak, the temperature and humidity control loop can still maintain good robust stability.

[0122] Overall, this solution avoids decoupling distortion and control oscillation caused by increased system response inertia, significantly extends the replacement cycle of filter consumables, reduces workshop operation and maintenance costs, and ensures high-precision temperature and humidity environment control quality of the coating production line throughout the entire consumable replacement cycle.

[0123] Example 3:

[0124] like Figure 2 As shown, existing temperature and humidity control in painting workshops often employs single-loop proportional-integral-derivative feedback control, or adjusts by setting fixed feedforward decoupling parameters under specific operating conditions. However, when faced with the conditions of variable air volume adjustment and simultaneous large-scale changes in multi-station spraying load in the painting process, the system exhibits strong temperature and humidity cross-coupling effects.

[0125] Furthermore, the existing system relies heavily on exhaust sensors for passive feedback. Due to the significant interference from paint mist inside the exhaust duct, it is difficult to obtain the true load status in real time. This ultimately leads to frequent technical problems such as temperature and humidity decoupling failure, control response lag, and coupling oscillation under variable air volume and variable load conditions.

[0126] To address the aforementioned issues of physical equipment response lag and cross-interference, this embodiment provides a specific implementation plan that combines system hardware entities with software logic. This system can identify real-time loads based on offline calibrated parameter models and online-acquired production actions, thereby dynamically offsetting the cross-interference effects of surface cooling and steam humidification during actual operation and suppressing the temperature-humidity coupling oscillation phenomenon that easily occurs during operating condition switching.

[0127] Specifically, this embodiment provides a state machine and a variable parameter temperature and humidity decoupling system for a coating workshop, which is applied to a control system for a coating workshop equipped with a variable air volume air conditioning unit. The control system is connected to an actuator valve group and includes: an offline calibration and modeling module, an online load identification module, a state machine scheduling module, an online parameter reconstruction module, and a decoupling compensation control module.

[0128] In actual industrial applications, the control system typically uses an industrial-grade programmable logic controller (PLC) or distributed control system (DCS) as the core edge computing control unit. The lower-level control cabinet communicates with the host computer (industrial PC and SCADA configuration system) in the central control room via an industrial Ethernet bus (such as PROFINET, EtherCAT, or ModbusTCP). The variable air volume (VAV) air conditioning unit includes high-power supply fans and exhaust fans, both equipped with variable frequency drives (VFDs).

[0129] The actuator valve group is located in the physical piping section of the air handling unit (AHU), specifically including a proportional regulating chilled water valve and a proportional regulating steam humidification valve equipped with an electrical intelligent positioner. It receives a 4-20mA DC current or 0-10V DC voltage analog control signal from the control system to adjust the actual valve core opening.

[0130] The system includes an offline calibration and modeling module, which is used to acquire multiple sets of offline air volume and spraying wet load operating points, identify the air thermodynamic four-channel transfer parameters under each operating point through the step response method, construct a parameter mapping model with relative air volume and relative total load as independent variables, and generate a control strategy library.

[0131] In terms of the coordinated implementation of physical hardware and software, this module is usually deployed as an independent engineering auxiliary component in the engineering station software of the host computer. During the no-load commissioning phase after the completion of the painting workshop construction, the engineering station sends a step-opening command to the electric positioners of the cold water valve and humidification valve through the communication bus, and simultaneously uses a high-speed data acquisition card to collect the dynamic response curve data of the high-precision platinum resistance temperature sensor (such as PT100 / PT1000 type) and polymer capacitive absolute humidity sensor installed at the end of the main air supply duct.

[0132] After being analyzed by the matrix fitting algorithm inside the software, the generated bivariate quadratic polynomial coefficient matrix (i.e., the control strategy library) will be downloaded via the underlying bus protocol and stored in the non-volatile memory register (such as FlashROM or NVRAM) of the underlying PLC controller, providing static basic mathematical model support for online control in the production process.

[0133] The system includes an online load identification module, which is used to collect the spray gun trigger signals of each spraying station and the volatile concentration data of the exhaust main pipe, and to identify and calculate the preliminary wet load of each spraying station and the real-time total spraying wet load of the workshop online through a soft measurement algorithm, and to calculate the real-time relative total load.

[0134] Because conventional physical humidity probes are prone to passivation of active coatings and clogging of dust covers in exhaust ducts containing high concentrations of paint mist and organic solvents, this module has undergone a redesign in its hardware data acquisition configuration. The control system relies on distributed remote I / O modules to acquire, in real time, 24VDC dry contact digital trigger signals output from the on-site robotic arm control cabinet or the pneumatic control valve of the manual spray gun via hardwiring.

[0135] Simultaneously, an online volatile organic compound (VOC) concentration monitor based on the principles of photoionization detector (PID) or flame ionization detector (FID) is installed outside the explosion-proof zone of the exhaust main duct or using an extraction sampling system to obtain an analog signal reflecting the evaporation rate of the coating. The CPU inside the PLC runs a recursive least squares soft measurement algorithm logic block with a forgetting factor in its main loop scan task, based on a hysteresis-free digital trigger signal and a highly reliable VOC analog signal.

[0136] This device is designed to achieve high-frequency, accurate online identification of the actual wet load of spraying on site without relying on easily damaged exhaust air physical humidity sensors.

[0137] The system includes a state machine scheduling module, which is used to input scheduling instructions, real-time air supply volume, duct static pressure and the real-time total wet spraying load of the workshop as four-dimensional trigger conditions into a multi-dimensional finite state machine. When it is determined that a state cross-interval switch has occurred, the corresponding reference parameter set is output. When it is fluctuating within the same interval, stepless parameter interpolation is performed.

[0138] In the program architecture of the underlying controller, this module is typically instantiated using Sequential Function Chart (SFC) or Structured Text (ST) language conforming to the IEC 61131-3 international standard. Its data acquisition hardware interfaces include: an Ethernet communication port for receiving production scheduling work orders from the upper-level Manufacturing Execution System (MES); a fieldbus interface for acquiring the actual operating frequency output of the blower inverter; and a thin-film differential pressure transmitter installed in the micro-positive pressure area of ​​the air supply system (for real-time acquisition of the pneumatic and static pressure of the duct).

[0139] This module acts as the core of the logic arbitration for the entire environmental control system. It uses a rigorous logical AND instruction chain to determine whether the four-dimensional physical and information conditions simultaneously and continuously meet the set anti-jitter time window. When the system confirms a substantial shift in the workshop operating conditions (such as switching from standby to spraying), it points the register addressing pointer to the reference parameter memory block for the corresponding process state. During periods of stable operating conditions, it uses the floating-point arithmetic unit to perform continuous linear or spline interpolation to prevent logical oscillations and actuator malfunctions during the transition phase of operating conditions.

[0140] The system includes an online parameter reconstruction module, which uses real-time relative air volume and real-time relative total load as dual scheduling variables, calls the parameter mapping model, and reconstructs online the temperature and humidity cross-feedforward decoupling coefficient and the operating parameters of the main loop controller. The main loop controller includes a temperature loop and a humidity loop.

[0141] This module uses the advanced floating-point arithmetic logic unit (ALU) inside the PLC as its calculation engine, and is synchronously triggered to run within each millisecond-level scan cycle. It takes the real-time relative total load output by the online load identification module and the real-time relative air volume calculated based on the frequency feedback of the frequency converter as real-time independent variable input values, and quickly substitutes them into the parameter mapping formula register downloaded by the offline calibration and modeling module.

[0142] Because the cooling and humidification sections of the air conditioning units in the painting workshop have huge thermophysical inertia, the feedforward decoupling coefficient (i.e., the ratio of the cross-physical gain of temperature-induced humidification and humidity-induced temperature) and the main loop PID operating parameters (proportional band, integral time constant, and derivative time constant) calculated by the parameter online reconstruction module at high frequency can directly reflect the aerodynamic and thermodynamic boundary constraints under the current transient conditions, ensuring that the embedded parameters of the controller always accurately follow the time-varying drift of the physical environment characteristics.

[0143] The system includes a decoupling compensation control module, which performs decoupling compensation calculations by combining the reconstructed temperature and humidity cross-feedforward decoupling coefficients with the operating parameters to generate an opening control signal to control the chilled water valve and humidification valve in the actuator valve group. In the system's execution layer closed loop, this module instantiates two independent standard PID algorithm function blocks in software memory, corresponding to the temperature closed loop and the humidity closed loop, respectively.

[0144] The calculation process first uses the deviation between the measured electrical signal and the set value of the air supply temperature and humidity sensor, combined with the reconstructed variable parameters, to perform basic control law calculations and generate initial control quantities.

[0145] Subsequently, before sending the data to the analog output channel (AO module), the outputs of both are compensated by algebraic addition and multiplication through the cross node with the feedforward decoupling coefficient.

[0146] The final generated opening control signal is converted into a standard industrial electrical signal (4-20mA) by the digital-to-analog converter (D / A) chip inside the PLC module, and transmitted to the field via a shielded cable to directly drive the pneumatic diaphragm actuator or servo motor that controls the cold water valve and humidification valve.

[0147] Based on the actual on-site production and use process, the variable air volume coating workshop state machine and variable parameter temperature and humidity decoupling system described in this embodiment can sense load jumps in advance through the soft measurement module when the coating production line faces frequent start-up and shutdown, rapid switching of process cycle, or drastic adjustment of fresh air volume. It can also sort out the control logic flow by relying on the state machine and directly issue coordinated action commands to the execution valve group by utilizing the online reconstruction and decoupling compensation module.

[0148] The system mechanism, at both the hardware and software levels, enables the actuator valve group to operate synchronously with the actual air supply fluctuations and the workstation's operating rhythm, dynamically offsetting the physical cross-interference caused by surface cooling dehumidification or steam heating, significantly reducing the regulation instability caused by strong coupling of multiple variables, and ensuring the constant operation of temperature and humidity in the workshop production environment.

[0149] Example 4:

[0150] Corresponding to the above embodiments, the present invention also proposes an electronic device.

[0151] like Figure 3 The diagram shows a structural schematic of an electronic device according to the present invention. The electronic device 100 includes a processor 101 and a memory 103. The processor 101 and the memory 103 are connected, for example, via a bus 102. Optionally, the electronic device 100 may further include a transceiver 104. It should be noted that in practical applications, the transceiver 104 is not limited to one unit, and the structure of this electronic device 100 does not constitute a limitation on the embodiments of the present invention.

[0152] Processor 101 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in connection with this disclosure. Processor 101 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0153] Bus 102 may include a pathway for transmitting information between the aforementioned components. Bus 102 may be a PCI bus or an EISA bus, etc. Bus 102 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0154] The memory 103 stores a computer program corresponding to the variable air volume painting workshop state machine and variable parameter temperature and humidity decoupling method of the above embodiments of the present invention. The computer program is controlled and executed by the processor 101. The processor 101 executes the computer program stored in the memory 103 to implement the contents shown in the aforementioned method embodiments.

[0155] Among them, electronic devices 100 include, but are not limited to: mobile terminals such as laptops and PADs (tablet computers) and fixed terminals such as desktop computers. Figure 3 The electronic device 100 shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0156] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A state machine for a variable air volume painting workshop and a variable parameter temperature and humidity decoupling method, characterized in that, A control system for a painting workshop equipped with a variable air volume (VAV) air conditioning unit, wherein the control system is connected to an actuator valve group, characterized in that it includes: Multiple sets of offline air volume and spraying wet load operating points are obtained. The air thermodynamic four-channel transfer parameters under each operating point are identified by the step response method. A parameter mapping model with relative air volume and relative total load as independent variables is constructed to generate a control strategy library. Collect the spray gun trigger signals and volatile concentration data of each spraying station and the exhaust manifold. Use a soft measurement algorithm to identify and calculate the preliminary wet load of each spraying station and the real-time total spraying wet load of the workshop online, and calculate the real-time relative total load. The scheduling command, real-time air supply volume, duct static pressure and the real-time total wet spraying load of the workshop are used as four-dimensional triggering conditions to input into the multi-dimensional finite state machine. When it is determined that a state cross-interval switch occurs, the corresponding benchmark parameter set is output. When it is fluctuating in the same interval, stepless parameter interpolation is performed. Using real-time relative air volume and real-time relative total load as dual scheduling variables, the parameter mapping model is invoked to reconstruct the temperature and humidity cross-feedforward decoupling coefficient and the operating parameters of the main loop controller online. The main loop controller includes a temperature loop and a humidity loop. By combining the reconstructed temperature and humidity cross-feedforward decoupling coefficient with the operating parameters, a decoupling compensation calculation is performed to generate an opening control signal to control the cold water valve and humidification valve in the actuator valve group.

2. The method according to claim 1, characterized in that, The construction of the parameter mapping model with relative air volume and relative total load as independent variables includes: The air thermodynamic four-channel transfer parameters include the static gain of the temperature main channel, the static gain of the humidity main channel, the static gain of the temperature-induced humidity coupling channel, and the static gain of the humidity-induced temperature coupling channel. Using the corresponding parameters under rated air volume and rated wet load as benchmark values, a bivariate quadratic polynomial structure is used to fit the air thermodynamic four-channel transfer parameters. The general expression of the generated parameter mapping model is: In the formula, In order to achieve a relative air volume of And the relative total load is Calculated values ​​of the target thermodynamic parameters under the given conditions. The reference value is the value at the corresponding operating point. , , , , , All are dimensionless mapping coefficients generated through offline fitting.

3. The method according to claim 1, characterized in that, The online identification and calculation of the preliminary wet load of each spraying station and the total real-time spraying wet load of the workshop through a soft measurement algorithm includes: By combining the pre-calibrated single-gun rated wet production rate of each spraying station with the spray gun trigger signal, the preliminary wet load of each spraying station is calculated and the total preliminary wet load is generated. Using the volatile concentration data as the observation benchmark, the overall load identification correction coefficient is calculated using a recursive least squares algorithm with a forgetting factor; Multiply the total preliminary wet load by the overall load identification correction coefficient to obtain the corrected real-time total spraying wet load in the workshop; The real-time relative total load is generated by dividing the workshop's real-time total wet load by the workshop's rated total wet load baseline value.

4. The method according to claim 3, characterized in that, The method also includes a step of allocating air supply branches on demand: Calculate the proportion of the initial wet load of each of the spraying stations in the total initial wet load; according to the proportion, decompose the corrected real-time total spraying wet load of the workshop to obtain the real-time wet load of each of the spraying stations. Using the real-time wet load of each workstation as the allocation weight, the total air supply volume is allocated to each air supply branch to obtain the target air demand volume of each air supply branch.

5. The method according to claim 1, characterized in that, The multidimensional finite state machine includes five discrete process states: hibernation, transition, standby, spraying, and self-cleaning. The state cross-interval switching must be triggered simultaneously by the following four-dimensional triggering conditions: the scheduling instruction matches the target working condition, the real-time air supply volume is within the air volume range of the target working condition for a first preset time, the static pressure of the air duct fluctuates within the set static pressure error range, and the real-time total wet load of the workshop spraying is within the load range of the target working condition for a second preset time. Furthermore, the airflow exit threshold and load exit threshold set in the state cross-interval switching are respectively lower than the corresponding airflow entry threshold and load entry threshold, forming an anti-shake hysteresis interval.

6. The method according to claim 5, characterized in that, The operation of the multidimensional finite state machine also includes: Collect workpiece queue and scheduling plan data from the production line, and predict the direction of work condition switching using a sliding window pattern matching algorithm; When the output condition switching prediction signal is received, before all four-dimensional triggering conditions are met, the current reference parameter set is proportionally fed forward to the parameter setting value corresponding to the target condition.

7. The method according to claim 2, characterized in that, The online reconstructed temperature and humidity cross-feedforward decoupling coefficients include: The real-time relative air volume is used as And the real-time relative total load is taken as Substitute the parameters into the parameter mapping model to calculate the current static gain of the temperature main channel, the static gain of the humidity main channel, the static gain of the temperature-induced humidity coupling channel, and the static gain of the humidity-induced temperature coupling channel. The ratio of the static gain of the temperature-induced humidity coupling channel to the static gain of the humidity main channel is negatively taken to generate a temperature-induced humidity feedforward decoupling coefficient used to offset the interference of surface cooling, condensation and dehumidification. The ratio of the static gain of the humidified temperature coupling channel to the static gain of the temperature main channel is negatively taken to generate a humidified temperature feedforward decoupling coefficient used to offset the interference of steam humidification temperature rise.

8. The method according to claim 7, characterized in that, The process of combining the reconstructed temperature and humidity cross-feedforward decoupling coefficients with the operating parameters to perform decoupling compensation calculations and generate an opening control signal to control the chilled water valve and humidification valve in the actuator valve group includes: Based on the operating parameters, proportional-integral-differential operations are performed in the temperature loop and the humidity loop respectively to generate initial temperature control quantity and initial humidity control quantity; The product of the initial temperature control quantity and the temperature-humidification feedforward decoupling coefficient is used as the first compensation quantity and added to the initial humidity control quantity to generate a control signal for controlling the opening of the humidification valve. The product of the initial humidity control value and the humidity-induced temperature feedforward decoupling coefficient is used as a second compensation value and added to the initial temperature control value to generate a control signal for controlling the opening degree of the cold water valve.

9. The method according to claim 4, characterized in that, After obtaining the target air volume requirement for each of the aforementioned air supply branches, the following is also included: Collect the actual air volume of each air supply branch, and calculate the branch air supply balance index by combining the target required air volume with the set air volume error lower limit. Set a first air supply deviation threshold and a second air supply deviation threshold, wherein the first air supply deviation threshold is greater than the second air supply deviation threshold; When the branch air supply balance index is greater than the first air supply deviation threshold, for air supply branches where the actual air volume deviates from the target air volume within a preset upper deviation range, the pulsation frequency of the corresponding air valve is increased and the static pressure compensation angle is decreased; for air supply branches where the actual air volume deviates from the target air volume within a preset lower deviation range, the pulsation frequency of the air valve is decreased and the static pressure compensation angle is increased.

10. The method according to claim 1, characterized in that, The method further includes a load adaptive disturbance rejection control step: A dynamic early warning threshold is calculated based on the real-time relative total load, and the dynamic early warning threshold decreases proportionally as the real-time relative total load increases. When the temperature and humidity control deviation is less than the dynamic early warning threshold, the control parameters of the decoupling compensation operation remain unchanged; When the temperature and humidity control deviation is greater than or equal to the dynamic early warning threshold, the temperature and humidity cross-feedforward decoupling coefficient is multiplied by a first proportional coefficient greater than one to amplify it, and the proportional band in the operating parameters is multiplied by a second proportional coefficient less than one to narrow it. When the actuator valve group is in the saturation opening limit state, the state switching process of the multidimensional finite state machine is locked, and the output change rate of the opening control signal is limited to within the set safe change rate threshold.