Rapid adjustment system and method for pressure difference of purification area

Abstract

The invention discloses a quick adjustment system for the pressure difference of a purification area, which comprises at least one tuyere adjustment device, at least one pressure difference detection device and a central control system, and the central control system comprises an input module, a data processing and operation module and an output module, the input module receives target pressure difference values set by a user for a plurality of to-be-adjusted areas, and the data processing and operation module receives real-time pressure difference value data and leather bag opening degree data and calculates and generates a control instruction for adjusting the opening degree of each leather bag through a control algorithm based on the received target pressure difference values. And the output module sends a control instruction of the opening degree of each leather bag to a remote control module of the tuyere adjusting device and drives the inflator pump to act to enable the real-time pressure difference value to approach to the target pressure difference value. The tuyere adjusting device has the advantage that pressure difference adjusting can be automatically, rapidly and accurately carried out.

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 present application aims 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 airflow 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 instruction for adjusting the opening degree of each bladder is calculated and generated through a control algorithm, and the output module sends the control instruction 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 to set the equivalent opening degree of the fixed air valve arranged in the area to be adjusted accordingly.

[0005] The present application also discloses a rapid adjustment method for purifying area pressure difference, comprising the following steps: S1, deployment stage: detachably installing the air port adjustment device on the inside of the air pipe of the area to be adjusted, and arranging the pressure difference detection device in the area to be adjusted and the reference area respectively; S2, setting stage: inputting the target pressure difference value required by each area to be adjusted to the input module of the central control system and starting the system; S3, adjustment stage: the central control system calculates and issues control instructions through a 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; S4, data recording and fixed air valve setting stage: the central control system records and outputs the final bladder opening degree suggestion value of the to-be-adjusted region pressure difference in the balanced state, and the worker manually adjusts the equivalent opening degree of the fixed air valve in the to-be-adjusted region; S5, disassembly and verification stage: disassemble and remove the pressure difference detection device and perform normal operation of the system, verify whether the to-be-adjusted region pressure difference value meets the requirements by using the fixed micro-pressure difference meter, if it meets the requirements, the system can be normally operated, if it does not meet the requirements, repeat steps S1-S5 until the to-be-adjusted region pressure difference value meets the requirements.

[0006] Further, the step S3 comprises the following steps: S31, system identification: fine-tune the bladder opening degree of the bladder arranged in each to-be-adjusted region, record the pressure difference corresponding data of all to-be-adjusted regions, and the central control system actively detects and quantifies the coupling relationship between the air ports inside each air pipe and constructs a response matrix; S32, cooperative optimization: on the basis of the response matrix constructed in step S31, a set of globally optimal initial bladder opening degrees that make the predicted pressure difference of all to-be-adjusted regions closest to the target pressure difference value is calculated by using the particle swarm algorithm, the global pressure difference error is minimized as the target and the optimal initial opening degree combination is searched and obtained, and the initial state of the system is placed at the position closest to the equilibrium point; S33, closed-loop fine-tuning: on the basis of the real-time pressure difference value data change feedback, the PID control algorithm is used to finely adjust the bladder opening degree in the to-be-adjusted region, and after several fine adjustments, the real-time pressure difference value of each to-be-adjusted region is stabilized at the target value.

[0007] Further, the step S31 comprises the following steps: Step S311, establishing initial state: the central control system outputs control signals to the air port adjustment device through the output module, and sets the bladder opening degree value of all bladders to the reference opening degree, the system enters the stable initial working condition state, the pressure difference detection device continuously monitors the pressure difference of each to-be-adjusted region and converts the pressure difference into a digital signal, and the data processing and operation module obtains the pressure difference digital signal and records the initial pressure difference value vector P_initial of each to-be-adjusted region; Step S312, sequential excitation and data acquisition: the data processing and operation module starts to execute the pre-set excitation sequence, instructs the bladder opening degree of one to-be-adjusted region to increase by a fixed step ΔV and keeps the bladder opening degrees of the remaining to-be-adjusted regions unchanged, after waiting for a pre-set stable time, the airflow in the air pipe re-enters the balanced state, the data processing and operation module records the pressure difference value change amount of all to-be-adjusted regions at this time and instructs the bladder with the increased fixed step to return to the reference opening degree, and the above process is repeated and the remaining to-be-adjusted regions are sequentially excited and data acquisition is performed to obtain the pressure difference value change amount data of the remaining to-be-adjusted regions. Step S313, constructing a response matrix: all the collected data are arranged into an n*n response matrix G; the step S32 uses the aforementioned response matrix G and calculates the initial opening combination of the bladders that makes the pressure difference values of all the regions to be adjusted approach the target most quickly by using the particle swarm optimization algorithm, the predicted pressure difference value of the region to be adjusted is calculated by the following formula: P predicted= P initial+ G*(V-V base), wherein V represents the bladder opening vector, V base is the reference opening uniformly set in step S311, after the optimization problem is converted into a target function to be minimized, the mathematical expression of the target function is J =||P target-P predicted||², wherein P target is the target pressure difference value vector set in step S2 and the bladder opening needs to be within its physically feasible range, after the data processing and operation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the individual history and group history optimal positions, updates the particle swarm speed 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 controls the output module to decompose the bladder opening vector V optimal into independent control instructions for the tuyere adjusting device, the output module sends the control instructions to the remote control module, the remote control module controls and drives the air pump to inflate or deflate, and adjusts the bladders to the bladder opening specified by the bladder opening vector V optimal at one time, the pressure difference detection device continuously monitors the real-time pressure difference value data P current and the data processing and operation module obtains the real-time pressure difference value data, calculates the current pressure difference error value e(t)=P target-P current for all the regions to be adjusted, the data processing and operation module starts the PID control algorithm, calculates the fine adjustment amount ΔU of the bladder opening according to the proportional, integral and differential items of the pressure difference error, the PID operation is fine-tuned on the basis of the globally optimal solution V optimal, i.e. U final=V optimal+ΔU, the output module sends the control instructions to the remote control module, the remote control module controls and drives the air pump to inflate or deflate to adjust the small bladder opening, after the fine-tuning is completed, the real-time pressure difference value data of all the regions to be adjusted is continuously monitored, when the real-time pressure difference value data of all the regions to be adjusted is continuously and stably within the allowable error range of the target pressure difference value within a preset time period, it is determined that the adjustment is completed and the closed-loop fine-tuning stage is exited, otherwise the closed-loop fine-tuning stage is restarted.

[0008] Further, the data processing and operation module starts the particle swarm optimization algorithm to obtain the globally optimal bladder opening vector includes the following steps: Step 1, algorithm initialization: particle coding is performed for the position of each particle i, X_i represents a potential tuyere skin bag opening degree combination, for a system with n tuyeres to be adjusted, the particle position is an n-dimensional vector, which can be expressed as X_i = [x_i1, x_i2,..., x_in], wherein x_ij represents the opening degree value of the jth skin bag suggested by the particle and 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 allocated within its feasible solution space, the initial speed V_i is randomly set, the fitness value of each particle is calculated according to the target function J = ||P_target-P_predicted||² defined in step S32, wherein P_predicted = P_initial+G*(X_i-V_base), the smaller the fitness value J is, the closer the predicted pressure difference value of the skin bag opening degree combination represented by the particle to the target pressure difference value, the better the position of the particle, the individual historical optimal 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 optimal position Gbest; Step 2, iterative updating and optimization: the particle swarm optimization algorithm enters an iterative loop with a maximum number of iterations T_max, and performs the following operations in each iteration t, update the speed of each particle i 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)), wherein ω represents the inertia weight and is used to balance the global exploration and local development ability, c_1 and c_2 are learning factors, both of which are usually positive constants and represent the step size of the particle learning to the individual historical and group historical optimal positions, r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] and are used to introduce randomness in the search, each dimension component of the speed V_i(t) is limited within the maximum speed V_max to prevent the search step from being too large and diverging, the particle position X_i(t) is updated according to the updated speed X_i(t)= X_i(t-1)+V_i(t), after updating, check whether each dimension component of X_i(t) exceeds the feasible range of the skin bag opening degree, if each dimension component 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 optimal position Pbest_i of the particle X_i(t), if J(X_i(t))<J(Gbest), update the global historical optimal position Gbest=X_i(t); Step 3, termination and output: when the number of iterations reaches the preset T_max or the global optimal fitness value J(Gbest) no longer significantly improves in continuous multiple iterations, the algorithm terminates, and the final output of the algorithm, i.e. the global historical optimal position Gbest, is the global optimal skin opening degree vector V_optimal sought.

[0009] Further, the data processing and operation module starts the PID control algorithm to calculate the skin opening degree fine tuning amount ΔU, which includes the following steps: Step (1), error definition and sampling: for the kth to be adjusted area, in each control period t, the error value e_k(t)=P_{target,k}-P_{current,k}(t) of the real-time pressure difference value is calculated, where P_{target, k} is the target pressure difference value set by the user, and P_{current, k}(t) is the real-time pressure difference value fed back by the pressure difference detection device at time t. The system samples and calculates e_k(t) at a fixed control period to form a discrete error sequence e_k(1), e_k(2),..., e_k(t); Step (2), calculate the discrete PID control amount: the skin opening degree fine tuning amount ΔU_k(t) of the skin corresponding to room k is composed of the proportional term, the integral term and the differential 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 a proportional adjustment effect with the current error instantaneous value, the integral term accumulates and weights all historical error values, and the differential term adjusts based on the change trend of the current error, i.e. the error change rate, and has the effects of leading prediction and damping; Step (3), parameter setting combination: after the ΔU_k(t) calculated in step (2) is subjected to amplitude limiting processing, the data processing and operation module generates a fine tuning control instruction and sends the control instruction to the remote control module through the output module, and the remote control module controls and drives the air pump to adjust the small skin opening degree of inflation or exhaust; Step (4), final skin bag opening degree value synthesis: after sending the final skin bag opening degree instruction U_k(t)=V_{optimal, k}+DeltaU_k(t) to the kth blast hole adjustment device at time t, the pressure difference values of all regions to be adjusted are continuously monitored, when the pressure difference error values e_k(t) of all regions to be adjusted remain in the preset error range for consecutive N control periods, it is determined that the closed-loop fine adjustment is completed and exited, at this time, the final U_k(t) is the final skin bag opening degree suggestion value recorded and output, otherwise, the closed-loop fine adjustment stage is restarted.

[0010] The beneficial effects of the present application are: the present application as a whole carries out pressure difference debugging through the detachable skin bag cooperating with the pressure difference detection device and the central control system, without affecting the normal structure and operation of the region to be adjusted, in the adjustment process, the control strategy of the whole region to be adjusted is intelligently controlled through the hybrid algorithm logic, the adjustment time of several hours or even several days in the existing adjustment mode is shortened to tens of minutes, the adjustment efficiency is greatly improved, the influence of the long adjustment time of the pressure difference on the normal work of the whole region to be adjusted is avoided, the particle swarm algorithm is used to solve the wind port coupling interference problem in the adjustment process from the global perspective, and the PID algorithm is combined for multiple accurate fine adjustment, which can accurately control the pressure difference and has a control precision far exceeding the existing manual repeated adjustment mode, greatly reducing the requirement for the operation experience of the workers, and greatly improving the overall adjustment efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0011] Fig. 1 is the overall architecture schematic diagram of the purification region pressure difference rapid adjustment system of the present application.

[0012] Fig. 2 is the flow chart of the purification region pressure difference rapid adjustment method of the present application.

[0013] Fig. 3 is the specific algorithm flow chart of step S3 of the purification region pressure difference rapid adjustment method of the present application. DETAILED DESCRIPTION

[0014] In order to make the technical means, innovative features and functions of the present application easy to understand, the present application will be further described below.

[0015] The technical scheme of the application is used for the rapid adjustment system of the differential pressure of the purification area, comprising at least one air port adjustment device, at least one differential pressure detection device and a central control system, the air port adjustment device is detachably installed on the inner side 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 inner side of the air pipe, the inflation pump is communicated 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, and then the opening degree of the bladder, that is, the degree of obstruction of the bladder to the air flow of the air pipe is changed, the differential pressure detection device is arranged in the area to be adjusted, the differential pressure 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 differential pressure 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 differential pressure 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 differential pressure value set by the user for multiple areas to be adjusted, the data processing and operation module receives the real-time differential pressure value data and the opening degree data of the bladder, and based on the received target differential pressure value, the control instruction for adjusting the opening degree of each bladder is calculated and generated through the control algorithm, the output module sends the control instruction 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 differential pressure value approaches the target differential pressure 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 to set the equivalent opening degree of the fixed air valve arranged in the area to be adjusted.

[0016] As Figs. 1-3As shown, a preferred embodiment of the present application mainly realizes the soft and hard cooperation of the control algorithm and the mechanical device through the linkage of the central control system, the differential pressure detection device and the tuyere adjusting device. In actual application, the tuyere adjusting device is detachably installed on the inner side of the air pipe in the area to be adjusted, so that the skin bag can expand or shrink under the inflation and exhaust of the air inlet and outlet on the skin bag and the inflation pump, and finally change the air pressure in the current area to be adjusted. The differential pressure detection device is arranged in the area to be adjusted and the reference area, so as to directly obtain the real-time differential pressure of the area to be adjusted and the reference area, facilitate the central control system to accurately obtain the real-time differential pressure of the area to be adjusted and the reference area, input the target differential pressure value required by each area to be adjusted into the input module of the central control system, and start the system. In step S3, the data processing and operation module controls the output module to issue a control instruction, and then the tuyere adjusting device fine tunes the skin bag opening degree of the skin bag arranged in each area to be adjusted, records the differential pressure corresponding data of all areas to be adjusted, and the central control system actively detects and quantifies the coupling relationship between the tuyeres on the inner side of each air pipe and constructs a response matrix. On the basis of the constructed response matrix, a set of global optimal initial skin bag opening degrees is calculated by using the particle swarm algorithm, so that the predicted differential pressure of all areas to be adjusted is closest to the target differential pressure value. The global differential pressure error is minimized as the target and the optimal initial opening degree combination is searched and obtained. The initial state of the system is placed at the position closest to the equilibrium point, while the oscillation and convergence time of the subsequent PID fine tuning stage is greatly reduced. In this step, the control algorithm is a hybrid algorithm combining the particle swarm algorithm and the PID control algorithm. The control algorithm identifies and constructs a response matrix describing the coupling relationship between the tuyeres on the inner side of each air pipe, and uses the particle swarm algorithm for global optimization calculation to output a set of global optimal initial skin bag opening degree values. Then, the PID control algorithm is used for several times of closed loop fine tuning to finally stabilize the real-time differential pressure value of each area to be adjusted at the target value. On this basis, the central control system records and outputs the final skin bag opening degree value of the area to be adjusted under the balanced state of the differential pressure, and the worker manually adjusts the equivalent opening degree of the fixed air valve arranged in the area to be adjusted to complete the differential pressure adjustment. After the adjustment is completed, the differential pressure detection device needs to be removed and the system needs to be operated normally. The fixed micro differential pressure gauge is used to verify whether the differential pressure of the area to be adjusted meets the requirements. If it meets the requirements, the system can be normally operated. If it does not meet the requirements, steps S1-S5 are repeated until the differential pressure of the current area to be adjusted meets the requirements.

[0017] In the actual application process, step S3 of the present application is realized through the three-step strategy of system identification, cooperative optimization and closed loop fine tuning. The multi-tuyere coupling interference is quickly and accurately eliminated, and the differential pressure of all areas to be adjusted is stabilized at the target value. This step is the core step of the intelligent adjustment of the present application. In a preferred embodiment of the present application, step S31 includes the following steps: Step S311, establishing an initial state: the central control system outputs a control signal to the tuyere adjustment device through the output module, and sets the skin bag opening degree value of all skin bags to a reference opening degree. In actual application, the reference opening degree can be flexibly set according to the actual application requirements. In this embodiment, it can be set to 50% according to the real-time requirements. After the setting is completed, the system enters a stable initial working condition. The differential pressure detection device continuously monitors the differential pressure of each to-be-adjusted region and converts the differential pressure into a digital signal. The digital signal is transmitted to the data processing and operation module through a signal transmission mode such as a wireless network. The data processing and operation module obtains the differential pressure digital signal and records the initial differential pressure value vector P_initial of each to-be-adjusted region. Step S312, sequential excitation and data acquisition: the data processing and operation module starts to execute a preset excitation sequence. In actual application, taking three to-be-adjusted regions provided with three tuyeres as an example, the three tuyeres are identified as V1, V2 and V3, and the three to-be-adjusted regions are identified as R1, R2 and R3. The skin bag opening degree of one to-be-adjusted region is increased by a fixed step length ΔV, which is set to 5% in this embodiment, and the skin bag opening degrees of the remaining to-be-adjusted regions remain unchanged. After a preset waiting time of 30 seconds, the airflow in the air pipe reenters a balanced state. The waiting time of 30 seconds is to ensure that the data collected is a valid steady-state corresponding value, rather than a transient data fluctuation, so as to avoid the influence of data fluctuation on the subsequent change amount collection. The data processing and operation module records the differential pressure value change amount of all to-be-adjusted regions and instructs the skin bag after the fixed step length is increased to return to the reference opening degree of 50%. At this time, ΔP11, ΔP21 and ΔP31 are obtained, wherein ΔP11 represents the influence of V1 on R1, ΔP21 represents the influence of V1 on R2, and ΔP31 represents the influence of V1 on R3. The foregoing process is repeated, and the remaining two to-be-adjusted regions are sequentially excited and data collected to obtain the differential pressure value change amount data ΔP12, ΔP22, ΔP32 and ΔP13, ΔP23, ΔP33 of the remaining two to-be-adjusted regions. Step S313, constructing a response matrix: all collected data are arranged into an n*n response matrix G. The response matrix G directly quantifies the coupling interference degree between the skin bags on the inner side of the air pipe. Delta P_{ij} represents the influence of the unit opening degree change of the skin bag j on the differential pressure of the to-be-adjusted region i math G = \begin{bmatrix} \Delta P_{11}&\Delta P_{12}&\Delta P_{13}\\ \Delta P_{21}&\Delta P_{22}&\Delta P_{23}\\ ΔΡ 31 ΔΡ 32 ΔΡ 33 \end{bmatrix}.

[0018] In the above embodiment, the step S32 calculates the initial combination of the bladder opening degrees of all rooms that makes the predicted pressure difference values of the to-be-adjusted areas closest to the target values by using the response matrix G and through a particle swarm optimization algorithm, so as to reduce the subsequent adjustment shock. The predicted pressure difference value of the to-be-adjusted area is calculated by the following formula: P predicted = P initial + G * (V-V base), wherein V represents the bladder opening degree vector, V base is the reference opening degree vector uniformly set in the step S311, and in this embodiment, it can be set as [50%, 50%, 50%]. After the optimization problem is converted into a minimization objective function, the mathematical expression of the objective function J = ||P target-P predicted|| 2, wherein P target is the target pressure difference value vector set in the step S2, and the bladder opening degree needs to be within the conventional 20%-100% physical feasible range. After the data processing and operation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the individual history and group history optimal positions, updates the particle swarm speed and position, and performs iterative search optimization. After several iterations, the algorithm converges and outputs the globally optimal bladder opening degree vector V optimal = [V1 optimal, V2 optimal, V3 optimal]. The central control system controls the output module to decompose the bladder opening degree vector V optimal into independent control instructions for the air outlet adjustment device. The output module sends the control instructions to the remote control module, the remote control module controls and drives the air pump to inflate or deflate, and the bladder is once adjusted to the bladder opening degree specified by the bladder opening degree vector V optimal. After completion, the system enters the most close-to-equilibrium point working condition from the stable initial working condition.

[0019] 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.

[0020] 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: 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. Step 2, iterative update and optimization: the particle swarm optimization algorithm enters an iterative loop with a maximum number of iterations T_max, and in each iteration t, the following operations are performed: for each particle i, update its 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 the global exploration and local development capabilities, c_1 and c_2 are learning factors, both are usually positive constants and represent the step size of the particle learning to the individual historical and group historical optimal position respectively, r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] and are used to introduce randomness in the search, each dimension component of the velocity V_i(t) is limited within the maximum velocity V_max to prevent the search step from being too large and diverging, update the particle position X_i(t) = X_i(t-1) + V_i(t) according to the updated velocity, check whether each dimension component of X_i(t) exceeds the feasible range of the bellows opening after updating, if each dimension component of X_i(t) exceeds the feasible range, constrain it on 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 optimal position Pbest_i = X_i(t) of the particle, if J(X_i(t)) < J(Gbest), update the global historical optimal position Gbest = X_i(t); Step 3, termination and output: when the number of iterations reaches the preset T_max or the global optimal fitness value J(Gbest) no longer improves significantly in consecutive iterations, the algorithm terminates, and the final output of the algorithm is the global historical optimal position Gbest, which is the global optimal bellows opening vector V_optimal sought.

[0021] The core of the above particle swarm optimization algorithm lies in the calculation of the objective function J, which directly depends on the response matrix G established in step S31, the initial pressure difference value vector P_initial, and the reference opening vector V_base, ensuring a direct and close coupling relationship between the particle swarm optimization algorithm and the system identification phase. The ultimate goal of this algorithm is to find a set of optimal opening values of the bellows opening, so that the overall error between the system model prediction of all the regions to be adjusted and the target pressure difference value set by the user is minimized, solving the technical problem of strong coupling interference between the various tuyeres in the prior art, which makes it difficult to accurately balance the pressure difference, and the pressure difference balance is easily broken. In the output execution phase, the bellows opening vector V_optimal output by the particle swarm optimization algorithm will be immediately converted into specific control instructions by the output module of the central control system, driving the air pump to charge and discharge the bellows, completing the closed loop from digital optimization to physical adjustment.

[0022] Similarly, the data processing and operation module starts the PID control algorithm to calculate the fine adjustment amount ΔU of the bladder opening, which includes the following steps: Step (1), error definition and sampling: for the kth to be adjusted area, in each control period t, which can be set to 2 seconds in actual application, the error value e_k(t)=P_{target,k}-P_{current,k}(t) of the actual pressure difference value is calculated, wherein P_{target, k} is the target pressure difference value set by the user, and P_{current, k}(t) is the real-time pressure difference value fed back by the pressure difference detection device at time t. The system samples and calculates e_k(t) at a fixed control period to form a discrete error sequence e_k(1), e_k(2),..., e_k(t); Step (2), calculate the discrete PID control amount: the PID control algorithm for the skin of the room k corresponding to the opening degree of the fine adjustment amount ΔU_k(t) is composed of proportional, integral and differential items, 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)], wherein the proportional term produces a proportional adjustment to the current error instantaneous value, in the actual application environment, the area to be adjusted may deviate from the target due to some unexpected reasons, the proportional term can quickly respond and drive the skin to correct the skin opening degree by a large amplitude to make the pressure difference return to the target value, wherein 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 system's steady-state error, in the actual application environment, 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 accurate static balance, K_i as the integral coefficient determines the speed and ability of eliminating residual error, the differential term adjusts based on the change trend of the current error, i.e. the error rate, and has the functions of advance prediction and damping, when the pressure difference quickly recovers to the target value, the differential term sends a deceleration signal in advance to prevent the skin from adjusting too much 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 change trend, the aforementioned K_p, K_i, K_d three parameters are not fixed, in a preferred embodiment of the present application, the initial values of these parameters are determined by the particle swarm optimization algorithm in the step S32 of the optimization stage to find the global optimal skin opening degree vector V_optimal, so that the particle swarm optimization algorithm finds the optimal starting point, and also finds a set of optimal control parameters matched with the coupled PID controller to make the global system stable quickly; Step (3), parameter setting combination: after the ΔU_k(t) calculated in step (2) is limited, the data processing and operation module generates a fine adjustment control instruction, and the output module sends the control instruction to the remote control module, which controls and drives the air pump to adjust the small skin opening degree, in the actual application environment, the amplitude processing ensures that its absolute value does not exceed the preset maximum single fine adjustment amount, such as ±2% opening, to prevent single-step adjustment from causing oscillation; Step (4), final skin bag opening degree value synthesis: after sending the final skin bag opening degree instruction U_k(t)=V_{optimal,k}+ΔU_k(t) to the kth tuyere adjusting device at time t, the pressure difference values of all regions to be adjusted are continuously monitored, and when the pressure difference error values e_k(t) of all regions to be adjusted remain within the preset error range for N consecutive control periods, it is determined that the closed-loop fine adjustment is completed and exited, at this time the final U_k(t) is the final skin bag opening degree suggestion value recorded and output, otherwise the closed-loop fine adjustment stage is restarted.

Claims

1. A fast trim system for purging area differential pressure, characterized by: The application relates to a kind of air port adjustment devices, at least one differential pressure detection device and a central control system, the air port adjustment device is detachably installed in the air pipe inside the region to be adjusted, and the air port adjustment device includes 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, the bladder is fixedly provided with a suction cup adsorbed to the inside of the air pipe on one side, the inflation pump is communicated with the inner cavity of the bladder through an air inlet hole arranged on the side 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, and then the opening degree of the bladder, i.e. the degree of obstruction of the air pipe airflow, is changed, the differential pressure detection device is arranged in the region to be adjusted, the differential pressure detection device includes at least two air pressure sensors, the air pressure sensors are used to detect the static pressure of at least one region to be adjusted and a reference region and calculate the real-time differential pressure 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 differential pressure detection device, the central control system includes an input module, a data processing and operation module and an output module, the input module receives the target differential pressure value set by the user for multiple regions to be adjusted, the data processing and operation module receives the real-time differential pressure value data and the opening degree data of the bladder, and generates control instructions for adjusting the opening degree of each bladder based on the received target differential pressure value through a control algorithm, 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 differential pressure value approaches the target differential pressure 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 equivalent opening degree of the fixed air valve arranged in the region to be adjusted for the staff to set.

2. A method for fast trim using the fast trim system for purifying area differential pressure according to claim 1, characterized in that, The application comprises the following steps: S1, deployment stage: detachably install the air port adjustment device in the air pipe inside the region to be adjusted, and arrange the differential pressure detection device in the region to be adjusted and a reference region respectively; S2, setting stage: input the target differential pressure value required by each region to be adjusted to the input module of the central control system and start the system; S3, adjustment stage: the central control system calculates and issues control instructions through a control algorithm according to the real-time differential pressure value and the opening degree data of the bladder, finely adjusts the opening degree value of the bladder, forms a closed-loop control circuit, and stabilizes the real-time differential pressure value of each region at the target value; S4, data recording and fixed air valve setting stage: the central control system records and outputs the final opening degree suggestion value of the bladder when the differential pressure of the region to be adjusted is in a balanced state, and the staff manually adjusts the equivalent opening degree of the fixed air valve arranged in the region to be adjusted. S5, disassembly and verification stage: disassembly and differential pressure detection device and the normal operation of the system, using 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 normal operation, 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 method of claim 2, wherein: The step S3 comprises the following steps: S31, system identification: fine adjustment of the skin bag opening degree of the skin bag arranged in each area to be calibrated, recording the differential pressure corresponding data of all areas to be calibrated, the central control system actively detects and quantifies the coupling relationship between the tuyeres inside the air pipe and constructs a response matrix; S32, collaborative optimization: on the basis of the response matrix constructed in step S31, a set of global optimal initial skin bag opening degrees is calculated by using particle swarm algorithm to make the predicted differential pressure of all areas to be calibrated closest to the target differential pressure value, the global differential pressure error is minimized as the target and the optimal initial opening combination is searched and obtained, and the initial state of the system is placed in the position closest to the equilibrium point; S33, closed-loop fine tuning: on the basis of real-time differential pressure value data change feedback, PID control algorithm is used for fine adjustment of the skin bag opening degree in the area to be calibrated, and after several fine adjustments, the real-time differential pressure value of each area to be calibrated is stabilized at the target value.

4. The method of claim 3, wherein: The step S31 comprises the following steps: Step S311, establishing initial state: the central control system outputs control signals to the tuyere calibration device through the output module, and the skin bag opening degree value of all skin bags is uniformly set to the reference opening degree, the system enters the stable initial working condition, the differential pressure detection device continuously monitors the differential pressure of each area to be calibrated and converts the differential pressure into digital signal, and the data processing and operation module obtains the differential pressure digital signal and records the initial differential pressure value vector P_initial of each area to be calibrated; Step S312, sequential excitation and data acquisition: the data processing and operation module starts to execute the preset excitation sequence, instructs the skin bag opening degree of one area to be calibrated to increase a fixed step ΔV and keeps the skin bag opening degrees of the remaining areas to be calibrated unchanged, after waiting for a preset stable time, the airflow in the air pipe reenters the equilibrium state, the data processing and operation module records the differential pressure value change of all areas to be calibrated at this time and instructs the skin bag after increasing the fixed step to restore to the reference opening degree, the above process is repeated and the remaining areas to be calibrated are sequentially excited and data acquisition is performed to obtain the differential pressure value change data of the remaining areas to be calibrated; Step S313, constructing a response matrix: all the collected data are arranged into an n*n response matrix G, the step S32 calculates the initial opening combination of the bladder that makes the pressure difference value of all the to-be-adjusted regions closest to the target by using the aforementioned response matrix G and through a particle swarm optimization algorithm, the predicted pressure difference value of the to-be-adjusted region is calculated by the following formula: P_predicted = P_initial + G*(V-V_base), wherein V represents the bladder opening vector, V_base is the reference opening uniformly set in step S311, after the optimization problem is converted into a minimized objective function, the mathematical expression of the objective function is J = ||P_target-P_predicted||², wherein P_target is the target pressure difference value vector set in step S2 and the bladder opening needs to be within its physically feasible range, after the data processing and operation module starts the particle swarm optimization algorithm, the particle swarm flies in the solution space and tracks the individual history and group history optimal positions, updates the particle swarm speed 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 controls the output module to decompose the bladder opening vector V_optimal into independent control instructions for the tuyere adjustment device, the output module sends the control instructions to the remote control module, the remote control module controls and drives the air pump to inflate or exhaust, and the bladder is once synchronously adjusted to the bladder opening specified by the bladder opening vector V_optimal, the pressure difference detection device continuously monitors the real-time pressure difference value data P_current and the data processing and operation module obtains the real-time pressure difference value data, the current pressure difference error value e(t)=P_target-P_current is calculated for all the to-be-adjusted regions, the data processing and operation module starts the PID control algorithm, calculates the fine adjustment amount ΔU of the bladder opening according to the proportional, integral and differential items of the pressure difference error, the PID operation is fine-tuned on the basis of the globally optimal solution V_optimal, that is, U_final=V_optimal+ΔU, the output module sends the control instructions to the remote control module, the remote control module controls and drives the air pump to inflate or exhaust to adjust the small bladder opening, after the fine-tuning is completed, the real-time pressure difference value data of all the to-be-adjusted regions are continuously monitored, when the real-time pressure difference value data of all the to-be-adjusted regions continuously and stably fall within the target pressure difference value allowable error range within a preset time period, it is determined that the adjustment is completed and the closed-loop fine-tuning stage is exited, otherwise the closed-loop fine-tuning stage is restarted.

5. The method of claim 4, wherein: The data processing and operation module starts the particle swarm optimization algorithm to obtain the globally optimal bladder opening vector includes the following steps: Step 1, algorithm initialization: particle coding is performed for the position of each particle i, X_i represents a potential tuyere skin bag opening degree combination, for a system with n to be adjusted tuyere, the particle position is an n-dimensional vector, which can be expressed as X_i = [x_i1, x_i2,..., x_in], wherein x_ij represents the opening degree value of the jth skin bag suggested by the particle and 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 allocated within its feasible solution space, the initial speed V_i is randomly set, the fitness value of each particle is calculated according to the target function J = ||P_target-P_predicted||² defined in step S32, wherein P_predicted = P_initial+G*(X_i-V_base), the smaller the fitness value J is, the closer the predicted pressure difference value represented by the skin bag opening degree combination of the particle to the target pressure difference value, the better the position of the particle, the individual historical optimal position Pbest_i of each particle is initialized as its current position, that is, 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 optimal position Gbest; Step 2, iterative update and optimization: the particle swarm optimization algorithm enters an iterative loop with a maximum iteration number of T_max, and performs the following operations in each iteration t, update the speed of each particle i 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)), wherein ω represents the inertia weight and is used to balance the global exploration and local development ability, c_1 and c_2 are learning factors, both of which are usually positive constants and represent the step size of the particle learning to the individual historical and group historical optimal positions, r_1 and r_2 are random numbers uniformly distributed in the interval [0, 1] and are used to introduce randomness in the search, each dimension component of the speed V_i(t) is limited within the maximum speed V_max to prevent the search step from being too large and diverging, the particle position X_i(t) is updated according to the updated speed X_i(t) = X_i(t-1) + V_i(t), after updating, check whether each dimension component of X_i(t) exceeds the feasible range of the skin bag opening degree, if each dimension component of X_i(t) exceeds the feasible range, constrain it on 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 optimal position Pbest_i of the particle X_i(t), if J(X_i(t)) < J(Gbest), update the global historical optimal position Gbest = X_i(t); Step 3, termination and output: when the number of iterations reaches the preset T_max or the global optimal fitness value J(Gbest) no longer significantly improves in continuous iterations, the algorithm is terminated, and the final output, i.e., the global historical optimal position Gbest, is the global optimal skin opening degree vector V_optimal sought.

6. The method of claim 4 or 5, wherein: The data processing and operation module starts a PID control algorithm to calculate the skin opening degree fine tuning amount ΔU, including the following steps: Step (1), error definition and sampling: for the kth area to be adjusted, in each control period t, the error value e_k(t)=P_{target,k}-P_{current,k}(t) of the real-time pressure difference value is calculated, where P_{target, k} is the target pressure difference value set by the user, and P_{current, k}(t) is the real-time pressure difference value fed back by the pressure difference detection device at time t. The system samples and calculates e_k(t) at a fixed control period to form a discrete error sequence e_k(1), e_k(2),..., e_k(t); Step (2), calculation of discrete PID control amount: the skin opening degree fine tuning amount ΔU_k(t) of the skin corresponding to room k is composed of proportional, integral and derivative terms, 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 a proportional adjustment to the current error instantaneous value, the integral term accumulates and weights all historical error values, and the derivative term adjusts based on the change trend of the current error, i.e., the error change rate, and has the effects of leading prediction and damping; Step (3), parameter setting combination: after the ΔU_k(t) calculated in step (2) is subjected to amplitude limiting processing, the data processing and operation module generates a fine tuning control instruction and sends the control instruction to the remote control module through the output module, and the remote control module controls and drives the air pump to adjust the skin opening degree by a small amount of inflation or deflation; Step (4), final skin opening degree value synthesis: after sending the final skin opening degree instruction U_k(t)=V_{optimal, k}+ΔU_k(t) to the kth air port adjustment device at time t, the pressure difference values of all areas to be adjusted are continuously monitored, and when the pressure difference error values e_k(t) of all areas to be adjusted remain within the preset error range for continuous N control periods, it is determined that the closed-loop fine tuning is completed and exited, and at this time the final U_k(t) is the final skin opening degree suggestion value recorded and output, otherwise the closed-loop fine tuning phase is restarted.

Images