A quick trim system and method for purging area differential pressure

Abstract

The application discloses a kind of for purifying area pressure difference quick adjustment system, including at least one air port adjustment device, at least one pressure difference detection device and a central control system, central control system includes input module, data processing and operation module and output module, the input module receives target pressure difference value that user sets for multiple to be adjusted area, the data processing and operation module receives real-time pressure difference value data and skin bag opening degree data, and based on received target pressure difference value, control instruction for adjusting each skin bag opening degree is calculated and generated by control algorithm, the output module sends the control instruction of each skin bag opening degree to the remote control module of air port adjustment device and drives the action of air pump, so that real-time pressure difference value approaches target pressure difference value, the application has the advantage that can automatically and quickly accurately carry out pressure difference debugging.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of differential pressure adjustment in a region, and in particular to a quick adjustment system and method for differential pressure in a clean region such as an operating room in a hospital, a biological laboratory, etc. BACKGROUND

[0002] In a hospital operating department, a negative pressure isolation ward, and a biological safety laboratory, maintaining a stable pressure difference between a specific room region and an adjacent region is the core of ensuring environmental safety. In this process, positive pressure is used to prevent pollutants from entering the current region, while negative pressure is used to prevent harmful substances from leaking. Currently, the establishment and balance of the pressure difference in a clean region mainly relies on the constant air volume valve and the variable air volume valve in the ventilation and air conditioning system. The existing pressure difference adjustment technology is directly integrated in the fixed system, lacks a separate application of professional tools, and after the initial debugging, regular maintenance, or layout changes of the system, experienced professional technicians need to manually adjust the valves of each air supply outlet, air return outlet, and air exhaust outlet through repeated attempts and observation of the micro pressure difference meter. This process is time-consuming, affects the normal use of the current region, and relies too much on human experience. On the other hand, this adjustment method is strongly coupled and interfered between the air outlets, and adjusting a single valve will trigger a chain reaction in the pressure difference of other regions to be adjusted, making it difficult to accurately balance the pressure difference. The pressure difference balance is easily broken. SUMMARY

[0003] The purpose of the present application is to provide a quick adjustment system and method for the pressure difference in a clean region that can automatically and quickly and accurately adjust the pressure difference.

[0004] To solve the above technical problems, the present application adopts the following technical solutions: A rapid adjustment system for purifying area pressure difference, comprising at least one air port adjustment device, at least one pressure difference detection device and a central control system, the air port adjustment device is detachably installed on the inside of the air pipe of the area to be adjusted, the air port adjustment device comprises a bladder made of high-strength flexible material, an inflation pump integrated in the bladder, a remote control module and a wireless charging module arranged in the bladder, one side of the bladder is fixedly provided with a suction cup adsorbed on the inside of the air pipe, the inflation pump is in communication with the inner cavity of the bladder through the air inlet hole arranged on the side surface of the bladder, the wireless charging module charges the inflation pump and the remote control module, the remote control module receives external instructions and controls the inflation and exhaust actions of the inflation pump, the inflation and exhaust actions of the inflation pump are controlled to make the bladder expand or shrink, thereby changing the opening degree of the bladder, that is, the degree of obstruction of the bladder to the air flow of the air pipe, the pressure difference detection device is arranged in the area to be adjusted, the pressure difference detection device comprises at least two air pressure sensors, the air pressure sensors are used to detect the static pressure of at least one area to be adjusted and a reference area and calculate the real-time pressure difference value data between the two, which is uploaded to the central control system, the central control system is in communication connection with the air port adjustment device and the pressure difference detection device, the central control system comprises an input module, a data processing and operation module and an output module, the input module receives the target pressure difference value set by the user for multiple areas to be adjusted, the data processing and operation module receives the real-time pressure difference value data and the bladder opening degree data, and based on the received target pressure difference value, the control algorithm is used to calculate and generate control instructions for adjusting the opening degree of each bladder, and the output module sends the control instructions of the opening degree of each bladder to the remote control module of the air port adjustment device and drives the inflation pump to act so that the real-time pressure difference value approaches the target pressure difference value, after the automatic adjustment of the central control system is completed, the output module of the central control system outputs the final stable opening degree suggestion value of each air port adjustment device and provides the staff with the equivalent opening degree of the fixed air valve arranged in the area to be adjusted.

[0005] The present application also discloses a rapid adjustment method for purifying area pressure difference, comprising the following steps:

[0006] S1, deployment stage: the air port adjustment device is detachably installed on the inside of the air pipe of the area to be adjusted, and the pressure difference detection device is arranged in the area to be adjusted and a reference area respectively;

[0007] S2, setting stage: input the target pressure difference value required by each area to be adjusted to the input module of the central control system and start the system;

[0008] S3, adjustment stage: the central control system calculates and issues control instructions through the control algorithm according to the real-time collected pressure difference value and the bladder opening degree data, finely adjusts the opening degree value of the bladder, forms a closed loop control circuit, and stabilizes the real-time pressure difference value of each area at the target value;

[0009] S4. Data recording and fixed damper setting stage: The central control system records and outputs the final recommended value of the bladder opening when the pressure difference in the area to be adjusted is in a balanced state. The staff manually adjusts the equivalent opening of the fixed damper set in the area to be adjusted.

[0010] S5. Disassembly and Verification Phase: Disassemble the differential pressure detection device and run the system under normal conditions. Use a fixed micro differential pressure gauge to verify whether the differential pressure value of the area to be calibrated meets the requirements. If it meets the requirements, it can be run normally. If it does not meet the requirements, repeat steps S1-S5 until the differential pressure value of the current area to be calibrated meets the requirements.

[0011] Furthermore, step S3 includes the following steps:

[0012] S31. System identification: Fine-tune the opening of the bladders in each area to be calibrated, record the pressure difference response data of all areas to be calibrated, and the central control system actively detects and quantifies the coupling relationship between the air outlets inside each duct and constructs a response matrix.

[0013] S32. Collaborative optimization: Based on the response matrix constructed in step S31, the particle swarm optimization algorithm is used to calculate a set of globally optimal initial bladder openings that make the predicted pressure difference of all regions to be calibrated closest to the target pressure difference value. The goal is to minimize the global pressure difference error and search for the optimal combination of initial openings, placing the initial state of the system at the position closest to the equilibrium point.

[0014] S33. Closed-loop fine-tuning: Based on the feedback of real-time differential pressure data changes, a PID control algorithm is used to finely adjust the bladder opening in the area to be calibrated. After several fine-tunings, the real-time differential pressure value of each area to be calibrated stabilizes at the target value.

[0015] Furthermore, step S31 includes the following steps:

[0016] Step S311: Establishing the initial state: The central control system outputs a control signal to the air outlet adjustment device through the output module, and sets the opening value of all the air bladders to the reference opening value. The system enters a stable initial working condition state. The differential pressure detection device continuously monitors the differential pressure of each area to be adjusted and converts the differential pressure into a digital signal. The data processing and calculation module acquires the differential pressure digital signal and records the initial differential pressure value vector P_initial of each area to be adjusted.

[0017] Step S312, Sequential Excitation and Data Acquisition: The data processing and calculation module starts executing the preset excitation sequence, instructing the bladder opening of one area to be adjusted to increase by a fixed step size ΔV while keeping the bladder opening of all remaining areas to be adjusted unchanged. After a preset stabilization time, the airflow in the duct re-enters a balanced state. The data processing and calculation module records the change in pressure difference value of all areas to be adjusted at this time and instructs the bladder after the fixed step size to return to the reference opening. The above process is repeated, and the remaining areas to be adjusted are individually excited and data acquired in sequence to obtain the change in pressure difference value of all remaining areas to be adjusted.

[0018] Step S313: Construct the response matrix: Organize all collected data into an n*n response matrix G; Step S32 uses the aforementioned response matrix G and a particle swarm optimization algorithm to calculate the initial sac opening combination that makes the pressure difference value of all areas to be calibrated approach the target as quickly as possible. The predicted pressure difference value of the areas to be calibrated is calculated by the following formula: P_predicted = P_initial + G*(V - V_base), where V represents the sac opening vector, and V_base is the baseline opening uniformly set in step S311. After transforming the optimization problem into minimizing the objective function, its mathematical expression objective function J is... =||P_target-P_predicted||², where P_target is the target differential pressure vector set in step S2, and the bladder opening must be within its physical feasible range. After the data processing and calculation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the optimal positions of individual history and group history, updates the particle swarm velocity and position, and performs iterative search for optimization. After several iterations, the algorithm converges and outputs the globally optimal bladder opening vector V_optimal=[V1_opt,V2_opt,V3_opt]. The central control system control output module decomposes the bladder opening vector V_optimal into independent control commands for the air outlet adjustment device. The output module sends the control commands to the remote control module. The remote control module controls and drives the air pump to inflate or deflate, and synchronously adjusts the bladder to the bladder opening specified by the bladder opening vector V_optimal in one go. The differential pressure detection device The system continuously monitors the real-time differential pressure data P_current and obtains the real-time differential pressure data through the data processing and calculation module. For all areas to be calibrated, the current differential pressure error value e(t) = P_target - P_current is calculated. The data processing and calculation module starts the PID control algorithm and calculates the fine adjustment amount ΔU of the bladder opening based on the proportional, integral, and derivative terms of the differential pressure error. The PID calculation is fine-tuned based on the global optimal solution V_optimal, i.e., U_final = V_optimal + ΔU. The output module sends the control command to the remote control module, which controls and drives the air pump to adjust the inflation or deflation of the bladder opening. After the fine adjustment is completed, the system continuously monitors the real-time differential pressure data of all areas to be calibrated. When the real-time differential pressure data of all areas to be calibrated remains stable within the allowable error range of the target differential pressure value within a preset time period, the calibration is determined to be completed and the closed-loop fine-tuning stage is exited. Otherwise, the closed-loop fine-tuning stage is restarted.

[0019] Furthermore, the data processing and computation module initiates the particle swarm optimization algorithm to obtain the globally optimal skin opening vector, including the following steps:

[0020] Step 1, Algorithm Initialization: For each particle i, particle encoding is performed. X_i represents a potential combination of duct opening degrees. For a system with n ducts to be adjusted, the particle position is an n-dimensional vector, which can be represented as X_i = [x_i1, x_i2, ..., x_in], where x_ij represents the opening value of the j-th duct suggested by the particle, and its value is constrained within the physically feasible range. A population containing a preset population size of M particles is randomly generated. The initial position of each particle is randomly assigned within its feasible solution space, and the initial velocity V_i is randomly set. The fitness value of each particle is calculated according to the objective function J = ||P_target-P_predicted||² defined in step S32, where P_predicted = P_initial+G*(X_i-V_base). The smaller the value, the closer the predicted pressure difference value of the particle is to the target pressure difference value. The better the position of the particle, the better the individual historical best position Pbest_i of each particle is initialized as its current position, i.e., Pbest_i=X_i. The particle with the smallest fitness value J in the population is found and its position is recorded as the global historical best position Gbest.

[0021] Step 2, Iterative Update and Optimization: The particle swarm optimization algorithm enters an iterative loop with a maximum number of iterations T_max, and performs the following operation in each iteration t: update the velocity V_i(t) of each particle i = ω * V_i(t-1) + c_1 * r_1 * (Pbest_i - X_i(t-1)) + c_2 * r_2 * (Gbest - X_i(t-1)), where ω represents the inertia weight and is used to balance global exploration and local exploitation capabilities, c_1 and c_2 are learning factors, which are usually positive constants and represent the step size for the particle to learn from the optimal position in the individual history and the group history, respectively, and r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] to introduce randomness into the search. Each component of the velocity V_i(t) is restricted to the maximum velocity V_max to prevent the search step size from being too large and diverging. Update the particle position X_i(t) according to the updated velocity = After updating X_i(t-1)+V_i(t), check whether each dimension of X_i(t) exceeds the feasible range of the sac opening. If each dimension of X_i(t) exceeds the feasible range, constrain it to the boundary value. Calculate the fitness value J of each particle at the new position X_i(t). If J(X_i(t))<J(Pbest_i), update the individual historical best position Pbest_i=X_i(t). If J(X_i(t))<J(Gbest), update the global historical best position Gbest=X_i(t).

[0022] Step 3, Termination and Output: The algorithm terminates when the number of iterations reaches the preset T_max or the global optimal fitness value J (Gbest) no longer shows significant improvement in consecutive iterations. The final output of the algorithm is the global historical optimal position Gbest, which is the globally optimal skin opening vector V_optimal.

[0023] Furthermore, the data processing and calculation module initiates the PID control algorithm to calculate the fine adjustment amount ΔU of the skin opening, including the following steps:

[0024] Step (1), Error Definition and Sampling: For the k-th region to be calibrated, calculate the error value of its real-time differential pressure value e_k(t)=P_{target,k}-P_{current,k}(t) in each control cycle t, where P_{target, k} is the target differential pressure value set by the user, and P_{current, k}(t) is the real-time differential pressure value fed back by the differential pressure detection device at time t. The system samples and calculates e_k(t) in a fixed control cycle to form a discrete error sequence e_k(1), e_k(2),...,e_k(t);

[0025] Step (2), Calculate the discrete PID control quantity: For the bladder corresponding to room k, the PID control algorithm has a bladder opening fine adjustment quantity ΔU_k(t) composed of a proportional term, an integral term and a derivative term, ΔU_k(t) = K_{p,k} \cdot e_k(t) + K_{i,k} \cdot T \cdot \sum_{j=0}^{t} e_k(j) + \frac{K_{d,k}}{T} \cdot [e_k(t) - e_k(t-1)], where the proportional term produces an adjustment effect proportional to the current instantaneous error value, the integral term accumulates and weights all historical error values, and the derivative term adjusts based on the current error change trend, i.e., the error change rate, and has the functions of advance prediction and damping;

[0026] Step (3), parameter tuning combination: After the ΔU_k(t) calculated in step (2) is limited, the data processing and calculation module generates a fine-tuning control command and the output module sends the control command to the remote control module. The remote control module controls and drives the air pump to adjust the inflation or deflation of the small bladder opening.

[0027] Step (4) Final bladder opening value synthesis: After sending the final bladder opening command U_k(t)=V_{optimal, k}+ΔU_k(t) to the k-th air outlet adjustment device at time t, continuously monitor the differential pressure value of all areas to be adjusted. When the differential pressure error value e_k(t) of all areas to be adjusted remains within the preset error range for N consecutive control cycles, the closed-loop fine-tuning is determined to be completed and exited. At this time, the final U_k(t) is the recorded and output final bladder opening recommendation value. Otherwise, the closed-loop fine-tuning stage is restarted.

[0028] The beneficial effects of this invention are as follows: This invention utilizes a detachable bladder in conjunction with a differential pressure detection device and a central control system for differential pressure adjustment, without affecting the normal structure and operation of the area to be adjusted. During the adjustment process, a hybrid algorithm logic intelligently controls the overall control strategy of the area to be adjusted, reducing the adjustment time from several hours or even days in existing methods to tens of minutes, significantly improving adjustment efficiency and avoiding the impact of excessively long differential pressure adjustment time on the normal operation of the area to be adjusted. In terms of adjustment precision control, a particle swarm optimization algorithm is used to solve the problem of vent coupling interference during the adjustment process from a global perspective, combined with a PID algorithm for multiple precise fine-tuning. This allows for precise control of differential pressure while maintaining control accuracy far exceeding that of existing manual, repeated trial-and-error adjustment methods, significantly reducing the requirements for operator experience and greatly improving overall adjustment efficiency. Attached Figure Description

[0029] Fig. 1 This is a schematic diagram of the overall architecture of the rapid adjustment system for the pressure difference in the purification area according to the present invention.

[0030] Fig. 2 This is a flowchart of the rapid adjustment method for the pressure difference in the purification area according to the present invention.

[0031] Fig. 3 This is a flowchart illustrating the specific algorithm for step S3 of the rapid adjustment method for the pressure difference in the purification area according to the present invention. Detailed Implementation

[0032] To make the technical means, innovative features and functions of this invention easier to understand, the invention will be further described below.

[0033] This invention provides a technical solution for a rapid pressure differential adjustment system in a purified area, comprising at least one air outlet adjustment device, at least one pressure differential detection device, and a central control system. The air outlet adjustment device is detachably installed inside the ductwork of the area to be adjusted. The air outlet adjustment device includes a bladder made of high-strength flexible material, an air pump integrated inside the bladder, a remote control module, and a wireless charging module located inside the bladder. A suction cup fixed to the inside of the duct is fixed to one side of the bladder. The air pump communicates with the inner cavity of the bladder through an air inlet located on the side of the bladder. The wireless charging module charges the air pump and the remote control module. The remote control module receives external commands and controls the inflation and deflation of the air pump. By controlling the inflation and deflation of the air pump, the bladder expands or contracts, thereby changing the bladder opening, i.e., the degree to which the bladder obstructs the airflow in the ductwork. The pressure differential detection device is located within the area to be adjusted and includes at least two pressure sensors. The system is used to detect the static pressure of at least one area to be calibrated and a reference area, calculate the real-time pressure difference between them, and upload the data to the central control system. The central control system is communicatively connected to the air outlet calibration device and the pressure difference detection device. The central control system includes an input module, a data processing and calculation module, and an output module. The input module receives the target pressure difference value set by the user for multiple areas to be calibrated. The data processing and calculation module receives the real-time pressure difference value data and the bladder opening data, and calculates and generates control commands for adjusting the opening of each bladder based on the received target pressure difference value through a control algorithm. The output module sends the control commands for the opening of each bladder to the remote control module of the air outlet calibration device and drives the air pump to make the real-time pressure difference value approach the target pressure difference value. After the central control system completes automatic calibration, the output module of the central control system outputs the final stable bladder opening recommendation value of each air outlet calibration device, which is used by the staff to set the equivalent opening of the fixed air valves set in the area to be calibrated.

[0034] like Figs. 1-3As shown, a preferred embodiment of the present invention mainly achieves the hardware and software coordination of the control algorithm and mechanical device through the linkage of the central control system, the differential pressure detection device, and the air outlet adjustment device. In practical applications, the air outlet adjustment device needs to be detachably installed inside the air duct of the area to be adjusted, so that the air bladder can expand or contract under the air inlet and outlet of the air bladder and the air pump's inflation and deflation action, ultimately changing the air pressure in the area to be adjusted. The differential pressure detection device is set in the area to be adjusted and the reference area respectively to directly obtain the real-time pressure difference between the area to be adjusted and the reference area, which is convenient for the central control system. The system can promptly and accurately acquire the real-time pressure difference between the area to be calibrated and the reference area. It inputs the target pressure difference value required for each area to be calibrated into the input module of the central control system and starts the system. In step S3, the data processing and calculation module controls the output module to issue control commands, thereby fine-tuning the opening of the vents in each area to be calibrated by the vent adjustment device. It records the corresponding pressure difference data for all areas to be calibrated. The central control system actively detects and quantifies the coupling relationship between the vents inside each duct and constructs a response matrix. Based on the constructed response matrix, it uses a particle swarm optimization algorithm to calculate a set of predicted pressure differences for all areas to be calibrated. The goal is to find the globally optimal initial ventilator opening value that is closest to the target differential pressure value. This minimizes the global differential pressure error and searches for the optimal initial opening combination, placing the system's initial state closest to the equilibrium point while significantly reducing oscillation and convergence time in the subsequent PID fine-tuning stage. In this step, the control algorithm is a hybrid algorithm combining particle swarm optimization (PSO) and PID control. The control algorithm identifies and constructs a response matrix describing the coupling relationship between the air vents inside each duct, and uses PSO for global optimization calculations to output a set of globally optimal initial ventilator opening values. These values ​​are then applied through PID control. The algorithm performs several closed-loop fine-tuning steps to stabilize the real-time differential pressure values ​​of each area to be calibrated to the target value. Based on this, the central control system records and outputs the final bladder opening value when the differential pressure of the area to be calibrated is in a balanced state. The staff manually adjusts the equivalent opening of the fixed air valves set in the area to be calibrated to complete the differential pressure calibration. After calibration, the differential pressure detection device needs to be removed and the system needs to be put into normal operation. The fixed micro differential pressure gauge is used to verify whether the differential pressure of the area to be calibrated meets the requirements. If it meets the requirements, it can be operated normally. If it does not meet the requirements, steps S1-S5 are repeated until the differential pressure of the current area to be calibrated meets the requirements.

[0035] In practical applications, step S3 of this invention employs a three-step strategy of system identification, collaborative optimization, and closed-loop fine-tuning to quickly and accurately eliminate multi-vent coupling interference, stabilizing the pressure difference in all areas to be calibrated at the target value. This step is the core step for achieving intelligent calibration in this invention. In a preferred embodiment of this invention, step S31 includes the following steps:

[0036] Step S311: Establishing the initial state: The central control system outputs a control signal to the air outlet adjustment device through the output module, uniformly setting the air outlet opening value of all air bladders to the reference opening. In actual application, this reference opening can be flexibly set according to the actual application requirements. In this embodiment, it can be set to 50% according to real-time requirements. After the setting is completed, the system enters a stable initial working condition state. The differential pressure detection device continuously monitors the differential pressure of each area to be adjusted and converts the differential pressure into a digital signal. The digital signal is transmitted to the data processing and calculation module through signal transmission methods such as wireless network. The data processing and calculation module obtains the differential pressure digital signal and records the initial differential pressure value vector P_initial of each area to be adjusted.

[0037] Step S312, Sequential Excitation and Data Acquisition: The data processing and calculation module begins to execute the preset excitation sequence. In practical applications, taking three areas to be calibrated with three air vents as an example, the three air vents are labeled V1, V2, and V3, and the three areas to be calibrated are labeled R1, R2, and R3. The instruction is to increase the flap opening of one area to be calibrated by a fixed step size ΔV. In this embodiment, ΔV is set to 5%, and the flap opening of all remaining areas to be calibrated remains unchanged. After a preset 30-second waiting period, the airflow in the duct re-enters a balanced state. The preset 30-second stabilization time is to ensure that the acquired data is a valid steady-state response, not a static response. Instantaneous data fluctuations are avoided to prevent them from affecting subsequent data acquisition. The data processing and calculation module records the pressure difference changes in all areas to be calibrated at this time and instructs the bladder to recover to the reference opening of 50% after increasing the step size. At this time, ΔP11, ΔP21, and ΔP31 are obtained, where ΔP11 represents the effect of V1 on R1, ΔP21 represents the effect of V1 on R2, and ΔP31 represents the effect of V1 on R3. The above process is repeated, and the remaining two areas to be calibrated are individually excited and data is acquired to obtain the pressure difference change data ΔP12, ΔP22, ΔP32 and ΔP13, ΔP23, ΔP33 for the remaining two areas to be calibrated.

[0038] Step S313: Construct the response matrix: Organize all collected data into an n*n response matrix G. The response matrix G, as a digital model, directly quantifies the degree of coupling interference between the bladders inside the duct. Delta P_{ij} represents the effect of a unit opening change of bladder j on the pressure difference in the area i to be calibrated.

[0039] math

[0040] G = \begin{bmatrix}

[0041] \Delta P_{11}&\Delta P_{12}&\Delta P_{13}\\

[0042] \Delta P_{21}&\Delta P_{22}&\Delta P_{23}\\

[0043] \Delta P_{31}&\Delta P_{32}&\Delta P_{33}

[0044] \end{bmatrix}.

[0045] In the above embodiment, step S32 uses the aforementioned response matrix G and a particle swarm optimization algorithm to calculate the initial balloon opening combination that makes all room pressure differential values ​​approach the target as quickly as possible, reducing oscillations in subsequent adjustments. The predicted pressure differential value of the area to be adjusted is calculated by the following formula: P_predicted = P_initial + G*(V - V_base), where V represents the balloon opening vector, and V_base is the baseline opening vector uniformly set in step S311. In this embodiment, it can be set to [50%, 50%, 50%]. After transforming the optimization problem into minimizing the objective function, its mathematical expression objective function J = ||P_target-P_predicted||², where P_target is the target pressure difference vector set in step S2 and the bladder opening must be within its normal 20%-100% physical feasible range. After the data processing and calculation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the optimal position of individual history and group history, updates the particle swarm velocity and position and performs iterative search optimization. After several iterations, the algorithm converges and outputs the globally optimal bladder opening vector V_optimal=[V1_opt,V2_opt,V3_opt]. The central control system control output module decomposes the bladder opening vector V_optimal into independent control commands for the air outlet adjustment device. The output module sends the control commands to the remote control module. The remote control module controls and drives the air pump to inflate or deflate, and adjusts the bladder to the bladder opening specified by the bladder opening vector V_optimal in one go. After completion, the system enters the closest equilibrium point condition from the stable initial working condition.

[0046] After reaching the closest equilibrium point, rapid local fine-tuning of the system is required to address potential model errors or deviations caused by minor disturbances. The differential pressure detection device continuously monitors the real-time differential pressure data P_current, which is then acquired by the data processing and calculation module. For all areas to be calibrated, the current differential pressure error value e(t) = P_target - P_current is calculated. The data processing and calculation module then initiates a PID control algorithm, calculating the fine-tuning amount ΔU of the bladder opening based on the proportional, integral, and derivative terms of the differential pressure error. The PID calculation is fine-tuned based on the global optimal solution V_optimal, where U_final = V_optimal + ΔU. The output module sends control commands to the remote control module, which controls and drives the air pump to adjust the inflation or deflation of the bladder opening. After fine-tuning, the real-time differential pressure data of all areas to be calibrated is continuously monitored. In this embodiment, when the real-time differential pressure data of all areas to be calibrated remains stable within ±0.5 of the target differential pressure value allowable error for a preset 60 seconds... Within Pa, the calibration is determined to be complete and the closed-loop fine-tuning stage is exited; otherwise, the closed-loop fine-tuning stage is restarted. The preset time and allowable error value of this process can be selected according to the actual application situation to adapt to the current application scenario, rather than the predetermined value.

[0047] In the above embodiments, the data processing and computation module initiates the particle swarm optimization algorithm to obtain the globally optimal skin opening vector, including the following steps:

[0048] Step 1, Algorithm Initialization: For each particle i, particle encoding is performed. X_i represents a potential combination of duct opening degrees. For a system with n ducts to be adjusted, the particle position is an n-dimensional vector, which can be represented as X_i = [x_i1, x_i2, ..., x_in], where x_ij represents the opening value of the j-th duct suggested by the particle, and its value is constrained to the physically feasible range of 20% ≤ x_ij ≤ 100%. A population containing a preset population size of M particles is randomly generated, where M=50. The initial position of each particle is randomly assigned within its feasible solution space, and the initial velocity V_i is randomly set. The fitness value of each particle is calculated according to the objective function J = ||P_target-P_predicted||² defined in step S32, where P_predicted = P_initial+G*(X_i-V_base). The smaller the value, the closer the predicted pressure difference value of the particle is to the target pressure difference value. The better the position of the particle, the better the individual historical best position Pbest_i of each particle is initialized as its current position, i.e., Pbest_i=X_i. The particle with the smallest fitness value J in the population is found and its position is recorded as the global historical best position Gbest.

[0049] Step 2, Iterative Update and Optimization: The particle swarm optimization algorithm enters an iterative loop with a maximum number of iterations T_max, and performs the following operation in each iteration t: update the velocity V_i(t) of each particle i = ω * V_i(t-1) + c_1 * r_1 * (Pbest_i - X_i(t-1)) + c_2 * r_2 * (Gbest - X_i(t-1)), where ω represents the inertia weight and is used to balance global exploration and local exploitation capabilities, c_1 and c_2 are learning factors, which are usually positive constants and represent the step size for the particle to learn from the optimal position in the individual history and the group history, respectively, and r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] to introduce randomness into the search. Each component of the velocity V_i(t) is restricted to the maximum velocity V_max to prevent the search step size from being too large and diverging. Update the particle position X_i(t) according to the updated velocity = After updating X_i(t-1)+V_i(t), check whether each dimension of X_i(t) exceeds the feasible range of the sac opening. If each dimension of X_i(t) exceeds the feasible range, constrain it to the boundary value. Calculate the fitness value J of each particle at the new position X_i(t). If J(X_i(t))<J(Pbest_i), update the individual historical best position Pbest_i=X_i(t). If J(X_i(t))<J(Gbest), update the global historical best position Gbest=X_i(t).

[0050] Step 3, Termination and Output: The algorithm terminates when the number of iterations reaches the preset T_max or the global optimal fitness value J (Gbest) no longer shows significant improvement in consecutive iterations. The final output of the algorithm is the global historical optimal position Gbest, which is the globally optimal skin opening vector V_optimal.

[0051] The core of the aforementioned particle swarm optimization algorithm lies in the calculation of the objective function J. The calculation of the objective function J directly depends on the response matrix G, the initial pressure difference vector P_initial, and the baseline opening vector V_base established in step S31. This ensures the direct and close coupling between the particle swarm optimization algorithm and the system identification stage. The ultimate goal of this algorithm is to find a set of optimal opening values ​​for the air vents, minimizing the overall error between the pressure difference predicted by the system model for all areas to be calibrated and the target pressure difference value set by the user. This solves the technical problem in the existing technology where there is strong coupling interference between various air vents, making it difficult to achieve accurate pressure difference balance, and the pressure difference balance is easily broken. In the output execution stage, the air vent opening vector V_optimal output by the particle swarm optimization algorithm is immediately converted into specific control commands by the output module of the central control system, driving the air pump to inflate and deflate the air vents, completing the closed loop from digital optimization to physical adjustment.

[0052] Similarly, the data processing and calculation module initiates the PID control algorithm to calculate the fine adjustment amount ΔU of the skin opening, including the following steps:

[0053] Step (1), Error Definition and Sampling: For the k-th region to be calibrated, in each control cycle t, t can be set to 2 seconds in practical applications. The error value of the real-time differential pressure is calculated as e_k(t)=P_{target,k}-P_{current,k}(t), where P_{target, k} is the target differential pressure value set by the user, and P_{current, k}(t) is the real-time differential pressure value fed back by the differential pressure detection device at time t. The system samples and calculates e_k(t) with a fixed control cycle to form a discrete error sequence e_k(1), e_k(2),...,e_k(t);

[0054] Step (2), Calculate the discrete PID control quantity: For the bladder corresponding to room k, the PID control algorithm's bladder opening fine-tuning quantity ΔU_k(t) is composed of the proportional term, integral term, and derivative term superimposed, ΔU_k(t) = K_{p,k} \cdot e_k(t) + K_{i,k} \cdot T \cdot \sum_{j=0}^{t} e_k(j) + \frac{K_{d,k}}{T} \cdot [e_k(t)-e_k(t-1)], where the proportional term generates an adjustment effect proportional to the instantaneous value of the current error. In practical applications, the pressure difference in the area to be calibrated may suddenly deviate from the target due to some unexpected reasons. The proportional term can quickly respond and drive the bladder to make a large correction of the bladder opening, so that the pressure difference returns to the vicinity of the target. K_p, as the proportional coefficient, determines the strength of the system's response to the current error. The integral term accumulates and weights all historical error values ​​to eliminate the steady-state error of the system. In practical applications, if the pressure difference is slightly higher or lower than the target value for a long time, the integral term will gradually increase its output over time until the pressure difference is pushed back to the target value, achieving precise static balance. Ki, as the integral coefficient, represents the speed and ability to eliminate residual errors. The differential term is based on the change of the current error. The rate of change of the pressure difference is adjusted to control the trend of the error and has the functions of advance prediction and damping. When the pressure difference recovers rapidly toward the target value, the differential term issues a deceleration signal in advance to prevent the bladder from over-adjusting and causing overshoot and oscillation, thereby significantly improving the stability and response smoothness of the system. K_d, as the differential coefficient, determines the sensitivity of the system to the trend of change. The aforementioned three parameters K_p, K_i, and K_d are not fixed. In a preferred embodiment of the present invention, the initial values ​​of these parameters are determined by the particle swarm optimization algorithm in the collaborative optimization stage of step S32 when searching for the global optimal bladder opening vector V_optimal, so that the particle swarm optimization algorithm finds the optimal starting point and finds a set of optimal control parameters that match a set of mutually coupled PID controllers and can make the global system fast and stable.

[0055] Step (3), parameter tuning combination: After the ΔU_k(t) calculated in step (2) is limited, the data processing and calculation module generates a fine-tuning control command and the output module sends the control command to the remote control module. The remote control module controls and drives the air pump to adjust the inflation or deflation of the small bladder opening. In the actual application environment, the amplitude reduction processing ensures that its absolute value does not exceed the preset maximum single fine-tuning amount, such as ±2% opening, to prevent excessive single-step adjustment from causing oscillation.

[0056] Step (4) Final bladder opening value synthesis: After sending the final bladder opening command U_k(t)=V_{optimal,k}+ΔU_k(t) to the k-th air outlet adjustment device at time t, continuously monitor the differential pressure value of all areas to be adjusted. When the differential pressure error value e_k(t) of all areas to be adjusted remains within the preset error range for N consecutive control cycles, the closed-loop fine-tuning is determined to be completed and exited. At this time, the final U_k(t) is the recorded and output final bladder opening recommendation value. Otherwise, the closed-loop fine-tuning stage is restarted.

Claims

1. A rapid adjustment system for differential pressure in a purification area, characterized in that: The system includes at least one air outlet adjustment device, at least one differential pressure detection device, and a central control system. The air outlet adjustment device is detachably installed inside the ductwork in the area to be adjusted. The device comprises a bladder made of high-strength flexible material, an air pump integrated inside the bladder, a remote control module, and a wireless charging module located inside the bladder. A suction cup fixed to the inside of the duct is fixed to one side of the bladder. The air pump communicates with the inner cavity of the bladder through an air inlet on the side of the bladder. The wireless charging module charges the air pump and the remote control module. The remote control module receives external commands and controls the inflation and deflation of the air pump. By controlling the inflation and deflation of the air pump, the bladder expands or contracts, thereby changing the bladder opening, i.e., the degree to which the bladder obstructs the airflow in the ductwork. The differential pressure detection device is located within the area to be adjusted and includes at least two pressure sensors. These pressure sensors are used to detect at least one area to be adjusted. The static pressure of the domain and reference area is calculated, and the real-time pressure difference between the two is uploaded to the central control system. The central control system is communicatively connected to the air outlet calibration device and the pressure difference detection device. The central control system includes an input module, a data processing and calculation module, and an output module. The input module receives the target pressure difference value set by the user for multiple areas to be calibrated. The data processing and calculation module receives the real-time pressure difference value data and the bladder opening data, and calculates and generates control commands for adjusting the opening of each bladder based on the received target pressure difference value through a control algorithm. The output module sends the control commands for the opening of each bladder to the remote control module of the air outlet calibration device and drives the air pump to make the real-time pressure difference value approach the target pressure difference value. After the central control system completes automatic calibration, the output module of the central control system outputs the final stable bladder opening recommendation value of each air outlet calibration device, which is used by the staff to set the equivalent opening of the fixed air valves set in the areas to be calibrated.

2. A rapid calibration method using the rapid calibration system for pressure differential in a purification area as described in claim 1, characterized in that, Includes the following steps: S1. Deployment phase: The air outlet calibration device is detachably installed inside the air duct in the area to be calibrated, and the differential pressure detection device is set in the area to be calibrated and the reference area respectively; S2, Setting stage: Input the target differential pressure value required for each area to be calibrated into the input module of the central control system and start the system; S3, Calibration Phase: The central control system calculates and issues control commands through control algorithms based on the real-time collected differential pressure value and bladder opening data, finely adjusting the bladder opening value to form a closed-loop control loop until the real-time differential pressure value of each area stabilizes at the target value. S4. Data recording and fixed damper setting stage: The central control system records and outputs the final recommended value of the bladder opening when the pressure difference in the area to be adjusted is in a balanced state. The staff manually adjusts the equivalent opening of the fixed damper set in the area to be adjusted. S5. Disassembly and Verification Phase: Disassemble the differential pressure detection device and run the system under normal conditions. Use a fixed micro differential pressure gauge to verify whether the differential pressure value of the area to be calibrated meets the requirements. If it meets the requirements, it can be run normally. If it does not meet the requirements, repeat steps S1-S5 until the differential pressure value of the current area to be calibrated meets the requirements.

3. The rapid calibration method for a rapid calibration system for pressure differential in a purification area according to claim 2, characterized in that: Step S3 includes the following steps: S31. System identification: Fine-tune the opening of the bladders in each area to be calibrated, record the pressure difference response data of all areas to be calibrated, and the central control system actively detects and quantifies the coupling relationship between the air outlets inside each duct and constructs a response matrix. S32. Collaborative optimization: Based on the response matrix constructed in step S31, the particle swarm optimization algorithm is used to calculate a set of globally optimal initial bladder openings that make the predicted pressure difference of all regions to be calibrated closest to the target pressure difference value. The goal is to minimize the global pressure difference error and search for the optimal combination of initial openings, placing the initial state of the system at the position closest to the equilibrium point. S33. Closed-loop fine-tuning: Based on the feedback of real-time differential pressure data changes, a PID control algorithm is used to finely adjust the bladder opening in the area to be calibrated. After several fine-tunings, the real-time differential pressure value of each area to be calibrated stabilizes at the target value.

4. The rapid calibration method for a rapid calibration system for pressure differential in a purification area according to claim 3, characterized in that: Step S31 includes the following steps: Step S311: Establishing the initial state: The central control system outputs a control signal to the air outlet adjustment device through the output module, and sets the opening value of all the air bladders to the reference opening value. The system enters a stable initial working condition state. The differential pressure detection device continuously monitors the differential pressure of each area to be adjusted and converts the differential pressure into a digital signal. The data processing and calculation module acquires the differential pressure digital signal and records the initial differential pressure value vector P_initial of each area to be adjusted. Step S312, Sequential Excitation and Data Acquisition: The data processing and calculation module starts executing the preset excitation sequence, instructing the bladder opening of one area to be adjusted to increase by a fixed step size ΔV while keeping the bladder opening of all remaining areas to be adjusted unchanged. After a preset stabilization time, the airflow in the duct re-enters a balanced state. The data processing and calculation module records the change in pressure difference value of all areas to be adjusted at this time and instructs the bladder after the fixed step size to return to the reference opening. The above process is repeated, and the remaining areas to be adjusted are individually excited and data acquired in sequence to obtain the change in pressure difference value of all remaining areas to be adjusted. Step S313: Construct the response matrix: Organize all collected data into an n*n response matrix G. Step S32 uses the aforementioned response matrix G and a particle swarm optimization algorithm to calculate the initial sac opening combination that makes the pressure difference value of all areas to be calibrated approach the target as quickly as possible. The predicted pressure difference value of the areas to be calibrated is calculated by the following formula: P_predicted = P_initial + G*(V - V_base), where V represents the sac opening vector, and V_base is the baseline opening uniformly set in step S311. After transforming the optimization problem into minimizing the objective function, its mathematical expression objective function J = ||P_target-P_predicted||², where P_target is the target pressure difference vector set in step S2, and the sac opening must be within its physical feasible range. After the data processing and calculation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the optimal positions of individual and group histories. It updates the particle swarm velocity and position and performs iterative search for optimization. After several iterations, the algorithm converges and outputs the globally optimal sac opening vector V_optimal= [V1_opt, V2_opt, V3_opt], ​​the central control system's control output module decomposes the bladder opening vector V_optimal into independent control commands for the air vent adjustment device. The output module sends these control commands to the remote control module, which controls and drives the air pump to inflate or deflate the bladder, synchronously adjusting it to the bladder opening specified by the bladder opening vector V_optimal. The differential pressure detection device continuously monitors the real-time differential pressure data P_current, and the data processing and calculation module obtains this data. For all areas to be adjusted, the current differential pressure error value e(t) = P_target - P_current is calculated. The processing and calculation module initiates the PID control algorithm, calculates the fine adjustment amount ΔU of the bladder opening based on the proportional, integral, and derivative terms of the differential pressure error. The PID calculation is based on the global optimal solution V_optimal, i.e., U_final = V_optimal + ΔU. The output module sends the control command to the remote control module, which controls and drives the inflation pump to adjust the inflation or deflation of the bladder opening. After the fine adjustment is completed, the real-time differential pressure data of all areas to be adjusted is continuously monitored. When the real-time differential pressure data of all areas to be adjusted is continuously stable within the allowable error range of the target differential pressure value within a preset time period, the adjustment is determined to be completed and the closed-loop fine adjustment stage is exited; otherwise, the closed-loop fine adjustment stage is restarted.

5. The rapid calibration method for a rapid calibration system for pressure differential in a purification area according to claim 4, characterized in that: The data processing and computation module initiates the particle swarm optimization algorithm to obtain the globally optimal skin opening vector, including the following steps: Step 1, Algorithm Initialization: For each particle i, particle encoding is performed. X_i represents a potential combination of vent openings. For a system with n ducts to be adjusted, the particle position is an n-dimensional vector, which can be represented as X_i = [x_i1, x_i2, ..., x_in], where x_ij represents the opening value of the j-th vent suggested by the particle, and its value is constrained within the physically feasible range. A population containing a preset population size of M particles is randomly generated. The initial position of each particle is randomly assigned within its feasible solution space, and the initial velocity V_i is randomly set. The fitness value of each particle is calculated according to the objective function J = ||P_target-P_predicted||² defined in step S32, where P_predicted = P_initial+G*(X_i-V_base). The smaller the value, the closer the predicted pressure difference value of the particle is to the target pressure difference value. The better the position of the particle, the better the individual historical best position Pbest_i of each particle is initialized as its current position, i.e., Pbest_i=X_i. The particle with the smallest fitness value J in the population is found and its position is recorded as the global historical best position Gbest. Step 2, Iterative Update and Optimization: The particle swarm optimization algorithm enters an iterative loop with a maximum number of iterations T_max, and performs the following operation in each iteration t: update the velocity V_i(t) = ω*V_i(t-1) + c_1*r_1*(Pbest_i-X_i(t-1)) + c_2*r_2*(Gbest-X_i(t-1)), where ω represents the inertia weight and is used to balance global exploration and local exploitation capabilities; c_1 and c_2 are learning factors, which are usually positive constants and represent the step size for the particle to learn from the optimal position in the individual history and the group history, respectively; r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] and are used to introduce randomness into the search; each component of the velocity V_i(t) is restricted to the maximum velocity V_max to prevent the search step size from being too large and diverging; update the particle position X_i(t) according to the updated velocity = After updating X_i(t-1)+V_i(t), check whether each dimension of X_i(t) exceeds the feasible range of the sac opening. If each dimension of X_i(t) exceeds the feasible range, constrain it to the boundary value. Calculate the fitness value J of each particle at the new position X_i(t). If J(X_i(t))<J(Pbest_i), update the individual historical best position Pbest_i=X_i(t). If J(X_i(t))<J(Gbest), update the global historical best position Gbest=X_i(t). Step 3, Termination and Output: The algorithm terminates when the number of iterations reaches the preset T_max or the global optimal fitness value J (Gbest) no longer shows significant improvement in multiple consecutive iterations. The final output of the algorithm is the global historical optimal position Gbest, which is the globally optimal skin opening vector V_optimal.

6. A rapid calibration method for a rapid calibration system for pressure differential in a purification area according to claim 4 or 5, characterized in that: The data processing and calculation module initiates the PID control algorithm to calculate the fine adjustment amount ΔU of the bladder opening, including the following steps: Step (1), Error Definition and Sampling: For the k-th region to be calibrated, calculate the error value of its real-time differential pressure value e_k(t)=P_{target,k}-P_{current,k}(t) in each control cycle t, where P_{target, k} is the target differential pressure value set by the user, and P_{current, k}(t) is the real-time differential pressure value fed back by the differential pressure detection device at time t. The system samples and calculates e_k(t) in a fixed control cycle to form a discrete error sequence e_k(1), e_k(2),...,e_k(t); Step (2), Calculate the discrete PID control quantity: For the bladder corresponding to room k, the PID control algorithm has a bladder opening fine adjustment quantity ΔU_k(t) composed of a proportional term, an integral term and a derivative term, ΔU_k(t) = K_{p,k} \cdot e_k(t) + K_{i,k} \cdot T \cdot \sum_{j=0}^{t} e_k(j) + \frac{K_{d,k}}{T} \cdot [e_k(t)-e_k(t-1)], where the proportional term produces an adjustment effect proportional to the current instantaneous error value, the integral term accumulates and weights all historical error values, and the derivative term adjusts based on the current error change trend, i.e., the error change rate, and has the functions of advance prediction and damping; Step (3), parameter tuning combination: After the ΔU_k(t) calculated in step (2) is limited, the data processing and calculation module generates a fine-tuning control command and the output module sends the control command to the remote control module. The remote control module controls and drives the air pump to adjust the inflation or deflation of the small bladder opening. Step (4) Final bladder opening value synthesis: After sending the final bladder opening command U_k(t)=V_{optimal, k}+ΔU_k(t) to the k-th air outlet adjustment device at time t, continuously monitor the differential pressure value of all areas to be adjusted. When the differential pressure error value e_k(t) of all areas to be adjusted remains within the preset error range for N consecutive control cycles, the closed-loop fine-tuning is determined to be completed and exited. At this time, the final U_k(t) is the recorded and output final bladder opening recommendation value. Otherwise, the closed-loop fine-tuning stage is restarted.

Images