Blood flow control optimization method for extracorporeal circulation detection system

Abstract

The invention discloses a blood flow control optimization method for an extracorporeal circulation detection system, and belongs to the technical field of control optimization, and the method comprises the steps: constructing an extracorporeal circulation detection blood flow PID control system, constructing an improved builder optimization algorithm parameter setting module, and carrying out the optimization of the blood flow. An improved builder optimization algorithm is used for conducting iterative optimization on blood flow PID control parameters, optimal control parameters are output by evaluating a fitness function, the optimal control parameters are loaded to a blood flow PID controller module, and a blood flow execution adjusting module is driven to achieve quick response and high-precision steady-state tracking on target flow. According to the method, the improved builder optimization algorithm is used for setting the blood flow PID parameters, the double initialization, structure correction and reverse jump strategies are introduced, global search and local development are effectively balanced, premature convergence is avoided, the rising and adjusting time can be remarkably shortened, the optimization precision can be greatly improved, and the method is suitable for large-scale popularization and application. And rapid and stable control on the target blood flow is realized.

Description

Technical Field

[0001] This invention belongs to the technical field of control optimization, and particularly relates to a method for optimizing blood flow control in an extracorporeal circulation detection system. Background Technology

[0002] The main task of an extracorporeal circulation monitoring system is to construct highly realistic physiological or pathological circulatory conditions based on relevant national and industry standards, and to quantitatively evaluate the hydraulic performance, hemolytic characteristics, and biocompatibility of the tested equipment. In this process, blood flow rate, as the core excitation signal and controlled variable of the monitoring system, directly determines the validity of the detection data and the reliability of metrological traceability due to its control accuracy, stability, and dynamic response characteristics. If the flow control of the monitoring system itself has significant deviations or fluctuations, it will directly lead to distorted performance evaluation results of the tested equipment, and may even mislead the research and development direction or cause missed detection of unqualified products. However, constructing a high-precision extracorporeal circulation monitoring blood flow control system faces considerable engineering challenges. From a fluid dynamics perspective, the detection loop typically consists of a reservoir, temperature control unit, various flow resistance simulators, flow / pressure sensors, and a drive pump assembly, forming a complex system. Typical high-order, nonlinear, strongly coupled, and time-varying complex fluid dynamics systems present challenges. On one hand, the rheological properties of the test medium change nonlinearly with temperature and shear rate, causing the pipeline impedance characteristics to change constantly. On the other hand, to verify the performance of the device under test under extreme conditions, the testing scheme often requires rapid switching over a wide range from low-flow-rate pediatric mode to high-flow-rate adult mode, or simulation of the pulsating flow waveform of the human heart. These factors combined make it difficult for conventional linear control methods to simultaneously ensure the steady-state accuracy and dynamic performance of the system under all operating conditions. Therefore, a metaheuristic optimization algorithm with strong global optimization capabilities and efficient convergence characteristics is needed. Applying this algorithm to the tuning of the PID control parameters for blood flow in the testing system will undoubtedly help improve the automation level of the extracorporeal circulation testing system, reduce measurement uncertainty, and ensure the rigor of medical device testing. Summary of the Invention

[0003] To overcome the technical problems described in the background section, this invention provides a blood flow control optimization method for an extracorporeal circulation detection system. The method utilizes an improved builder optimization algorithm to tune the PID parameters of the blood flow. By introducing dual initialization, structural correction, and reverse jump strategies, it effectively balances global search and local development and avoids premature convergence. This significantly shortens the rise and adjustment time and greatly improves the optimization accuracy, thereby achieving rapid and stable control of the target blood flow.

[0004] The technical solution of the present invention is: a method for optimizing blood flow control in an extracorporeal circulation detection system, comprising the following steps: Step S1: Construct an extracorporeal circulation blood flow PID control system based on a closed-loop feedback mechanism, including a real-time blood flow monitoring module, a flow deviation calculation module, a blood flow PID controller module, a blood flow execution regulation module, and an improved builder optimization algorithm parameter tuning module. Step S2: Construct an improved builder optimization algorithm parameter tuning module. The improvements to the builder optimization algorithm specifically include: Step S21: Introduce a dual-scheme pre-selection initialization strategy. By performing random initial building scheme generation, reverse building scheme calculation, and hybrid screening based on fitness value, reverse solutions are introduced and selected for retention during the initialization phase, expanding the initial search range to complementary regions. Step S22: Introduce the master-led structural correction strategy. By setting a differential guided position update formula based on the top-level landmark blueprint and the reference blueprint during the exploration phase, the search direction for structural correction is determined by utilizing the guiding role of the global optimal solution and the random superior solution in the group. Step S23: Introduce a dynamic precision fine-tuning process strategy. By setting a construction precision radius calculation formula that decays exponentially with the construction progress and a local position update formula based on the construction precision radius during the development stage, the disturbance range is controlled to shrink nonlinearly with the number of iterations, thus realizing the transition from global breadth search to local depth development. Step S24: Introduce a reverse thinking design jump strategy. By setting up probability-triggered intergenerational jumps, calculating reverse design schemes for top-level landmark blueprints, and selectively replacing the worst individual in the population, a reverse perturbation for the global optimal solution is introduced and the worst individual is replaced. The complementary characteristics of the reverse solution are used to increase population diversity. Step S3: The improved builder optimization algorithm is used to tune the PID control parameters of the extracorporeal circulation blood flow detection blood flow PID control system. The optimal control parameters are output through iterative optimization. The optimal control parameters include the proportional coefficient. Integral coefficient and differential coefficients ; Step S4: The three optimal control parameters obtained using the improved builder optimization algorithm are loaded into the blood flow PID controller module of the extracorporeal circulation blood flow detection PID control system. The blood flow PID controller module executes PID control law calculations based on the optimal control parameters, and performs proportional, integral, and differential linear combination calculations on the real-time deviation between the set target blood flow and the monitored actual blood flow to generate a control signal. It also drives the blood flow regulation module through control signals to achieve rapid response and high-precision steady-state tracking of actual blood flow to target flow.

[0005] Furthermore, in the extracorporeal circulation blood flow detection PID control system constructed in step S1, the real-time blood flow monitoring module is used to collect the instantaneous blood flow signal in the extracorporeal circulation tubing and transmit it as a feedback quantity to the flow deviation calculation module. The flow deviation calculation module receives the set target blood flow and calculates the real-time deviation between the target value and the feedback quantity. The data is then transmitted to the blood flow PID controller module. The improved builder optimization algorithm parameter tuning module serves as the upper-level optimization unit, calculating the fitness function of the system through iterative optimization. Minimize the optimal parameter vector [ , , The data is then loaded into the blood flow PID controller module, which uses the received optimal parameters and real-time deviations. The control quantity is calculated based on the PID control law. It then outputs the control quantity to the blood flow control module, which responds to the control quantity. By changing the pumping state to adjust the actual blood flow, precise tracking and stable control of the target flow can be achieved.

[0006] Furthermore, the control law used in the blood flow PID controller module is: , In the formula Indicates the control output signal. This indicates the deviation between the set blood flow rate and the actual blood flow rate. This represents the proportionality coefficient. Represents the integral coefficient. This represents the differential coefficient.

[0007] Furthermore, the dual-scheme pre-selection initialization strategy in step S21 includes the following steps: Step S211: Generate a scale of within the search space. Random initial building scheme , , In the formula Indicates the lower bound of the search space. Indicates the upper bound of the search space. Represents a random vector between 0 and 1; Step S212: Based on random initial building scheme Calculate the corresponding reverse building scheme , , In the formula Indicates a reverse building scheme. Indicates the lower bound of the search space. Indicates the upper bound of the search space. This represents a random initial building scheme; Step S213: Randomize the initial building scheme Reverse building scheme Merging to form a scale of A set of mixed candidate solutions; Step S214: Using the fitness function Each solution in the mixed candidate solution set is evaluated to obtain the fitness value of each solution, and all solutions are sorted in ascending order of fitness value; Step S215: Select the items that are at the beginning after sorting. The proposed schemes were selected as the initial formal building population, and the scheme with the lowest fitness value was designated as the current top-level landmark blueprint. .

[0008] Furthermore, the master-led structural correction strategy in step S22 includes the following steps: Step S221: For each current scheme in the population Select all individuals from the population with fitness values ​​greater than [a certain value]. The proposed solutions constitute a set of superior solutions, and a solution is randomly selected from this set as a reference blueprint. ; Step S222: Utilize the top-level landmark blueprint in the current population. and the reference blueprint selected in step S221 Jointly guide the current plan The revised position update formula is as follows: , In the formula Indicates the location of the new scheme after structural modification. Indicates the location of the current building plan. Represents a random number between 0 and 1. This indicates that the top-level landmark blueprint represents the globally optimal solution. This refers to a random solution within the set of superior solutions, which is the reference blueprint.

[0009] Furthermore, the dynamic accuracy fine-tuning process strategy in step S23 includes the following steps: Step S231: Calculate the results as the number of iterations increases. Construction accuracy radius decreases exponentially , , In the formula This indicates the construction accuracy radius for the current iteration. This represents the initial radius scaling factor. Indicates the upper bound of the search space. Indicates the lower bound of the search space. Represents the precision convergence rate coefficient. Indicates the current iteration number. Indicates the maximum number of iterations; Step S232, Based on construction accuracy radius Apply isotropic random perturbations to the architectural design. , In the formula Indicates the location of the new design after fine polishing. Indicates the location of the current building plan. It represents a random vector between 0 and 1.

[0010] Furthermore, the reverse thinking design jump strategy in step S24 includes the following steps: Step S241: Generate a random number in each iteration and match the random number with the preset design jump probability. Compare the results to determine if the jump condition is met; Step S242: If the random number is less than The improved builder optimization algorithm calculates the current top-level landmark blueprint. Reverse design solutions within the search space , , In the formula This represents a reverse design scheme for a top-tier landmark blueprint. Indicates the lower bound of the search space. Indicates the upper bound of the search space. This represents the current blueprint for top-level landmarks; Step S243: Traverse the current building population and identify the solution with the highest fitness value as the discarded solution. ; Step S244: Calculate the reverse design scheme The fitness value, and compared with the discard scheme. Compare the fitness values, if The fitness value is less than The fitness value is then used. replace .

[0011] Furthermore, the improved builder optimization algorithm in step S3 uses a comprehensive performance index as the fitness function. , , In the formula Represents the fitness value. Indicates the simulation time. Indicates systematic error. This indicates the overshoot penalty weight. This indicates the adjustment of the time penalty weight. This indicates the actual overshoot. Indicates the target overshoot. Indicates the actual adjustment time. Indicates the target adjustment time.

[0012] The beneficial effects of the above technical solution are as follows: 1. This invention introduces a swarm intelligence optimization algorithm, namely an improved builder optimization algorithm. The blood flow PID controller tuned by the improved builder optimization algorithm has significantly shortened rise time and settling time of step response, and higher steady-state control accuracy. It can maintain the actual blood flow stably near the target value for a long time, reducing steady-state error. This makes the extracorporeal circulation detection system more adaptive and robust when dealing with extracorporeal circulation equipment or devices that simulate changes in the physiological state of patients or external disturbances. This greatly enhances the safety and reliability of equipment operation, helps to improve the automation level of the extracorporeal circulation detection system, helps to reduce measurement uncertainty, and helps to ensure the rigor of medical device testing.

[0013] 1. This invention expands the initial search range of the improved builder optimization algorithm to the complementary region of the candidate solution space by executing a dual scheme pre-selection initialization strategy that includes random initial building scheme generation and reverse building scheme calculation, as well as a hybrid screening process based on fitness values. It also increases the diversity of the initial building population by utilizing the complementary characteristics of the reverse solutions, thereby providing a high-quality initial population covering a wider solution space for the subsequent iterative optimization process. 2. This invention utilizes a master-led structural correction strategy guided by the difference between the top-level landmark blueprint and the reference blueprint through collaborative operation, and a dynamic precision fine-tuning process strategy that utilizes the construction precision radius that decays exponentially with the number of iterations. While using the global optimal solution and random superior solution in the group to determine the search direction, it controls the perturbation range to shrink nonlinearly with the number of iterations. This achieves a smooth transition from global breadth search to local depth development in the improved builder optimization algorithm and improves the local optimization accuracy in the later stages of iteration. 3. This invention performs probability-triggered generational jump and top-level landmark blueprint reverse design scheme calculation and optimal replacement process for the last individual in the population. During the iteration process, it introduces a reverse perturbation for the global optimal solution to explore a wider solution space and uses the complementary characteristics of the reverse solution to break the premature convergence state, thereby avoiding the algorithm from getting stuck in local optima and ensuring the quality of the existing optimal solution. 4. The present invention obtains the optimal scaling factor by optimizing the improved builder optimization algorithm. Optimal integral coefficient and optimal differential coefficients The PID controller module for blood flow is loaded to perform linear combination calculations for real-time flow deviations. This reduces the system fitness function value while shortening the rise time and settling time of the actual blood flow response, thus achieving rapid response and steady-state tracking of the target blood flow. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the present invention.

[0015] Figure 2 This is a curve comparing the optimal fitness values ​​of the builder optimization algorithm of this invention and the improved builder optimization algorithm.

[0016] Figure 3 This is a comparison curve of the step response between the builder optimization algorithm of this invention and the improved builder optimization algorithm.

[0017] Figure 4 This is a comparison curve of the PID parameter optimization process between the builder optimization algorithm of this invention and the improved builder optimization algorithm. Detailed Implementation

[0018] Example 1: As Figures 1-4 As shown, the present invention provides a method for optimizing blood flow control in an extracorporeal circulation detection system, comprising the following steps: Step S1: Construct a PID control system for detecting blood flow in extracorporeal circulation based on a closed-loop feedback mechanism; Step S2: Construct an improved builder optimization algorithm parameter tuning module. The improvements to the builder optimization algorithm specifically include: Step S21: Introduce a dual-scheme pre-selection initialization strategy. By performing random initial building scheme generation, reverse building scheme calculation, and hybrid screening based on fitness value, reverse solutions are introduced and selected for retention during the initialization phase, expanding the initial search range to complementary regions. Step S22: Introduce the master-led structural correction strategy. By setting a differential guided position update formula based on the top-level landmark blueprint and the reference blueprint during the exploration phase, the search direction for structural correction is determined by utilizing the guiding role of the global optimal solution and the random superior solution in the group. Step S23: Introduce a dynamic precision fine-tuning process strategy. By setting a construction precision radius calculation formula that decays exponentially with the construction progress and a local position update formula based on the construction precision radius during the development stage, the disturbance range is controlled to shrink nonlinearly with the number of iterations, thus realizing the transition from global breadth search to local depth development. Step S24: Introduce a reverse thinking design jump strategy. By setting up probability-triggered intergenerational jumps, calculating reverse design schemes for top-level landmark blueprints, and selectively replacing the worst individual in the population, a reverse perturbation for the global optimal solution is introduced and the worst individual is replaced. The complementary characteristics of the reverse solution are used to increase population diversity. Step S3: The improved builder optimization algorithm is used to tune the PID control parameters of the extracorporeal circulation blood flow detection blood flow PID control system. The optimal control parameters are output through iterative optimization. The optimal control parameters include the proportional coefficient. Integral coefficient and differential coefficients ; Step S4: The three optimal control parameters obtained using the improved builder optimization algorithm are loaded into the blood flow PID controller module of the extracorporeal circulation blood flow detection PID control system. The blood flow PID controller module executes PID control law calculations based on the optimal control parameters, and performs proportional, integral, and differential linear combination calculations on the real-time deviation between the set target blood flow and the monitored actual blood flow to generate a control signal. It also drives the blood flow regulation module through control signals to achieve rapid response and high-precision steady-state tracking of actual blood flow to target flow.

[0019] The extracorporeal circulation blood flow PID control system based on a closed-loop feedback mechanism includes a real-time blood flow monitoring module, a flow deviation calculation module, a blood flow PID controller module, a blood flow execution adjustment module, and an improved builder optimization algorithm parameter tuning module. The real-time blood flow monitoring module collects the instantaneous blood flow signal in the extracorporeal circulation tubing and transmits it as feedback to the flow deviation calculation module. The flow deviation calculation module receives the set target blood flow and calculates the real-time deviation between the target value and the feedback. The data is then transmitted to the blood flow PID controller module. The improved builder optimization algorithm parameter tuning module serves as the upper-level optimization unit, calculating the fitness function of the system through iterative optimization. Minimize the optimal parameter vector [ , , The data is then loaded into the blood flow PID controller module, which uses the received optimal parameters and real-time deviations. The control quantity is calculated based on the PID control law. It then outputs the control quantity to the blood flow control module, which responds to the control quantity. By changing the pumping state to adjust the actual blood flow, precise tracking and stable control of the target flow can be achieved.

[0020] The dual-scheme pre-selection initialization strategy in step S21 includes the following steps: Step S211: Generate a scale of within the search space. Random initial building scheme , , In the formula Indicates the lower bound of the search space. Indicates the upper bound of the search space. Represents a random vector between 0 and 1; Step S212: Based on random initial building scheme Calculate the corresponding reverse building scheme , , In the formula Indicates a reverse building scheme. Indicates the lower bound of the search space. Indicates the upper bound of the search space. This represents a random initial building scheme; Step S213: Randomize the initial building scheme Reverse building scheme Merging to form a scale of A set of mixed candidate solutions; Step S214: Using the fitness function Each solution in the mixed candidate solution set is evaluated to obtain the fitness value of each solution, and all solutions are sorted in ascending order of fitness value; Step S215: Select the items that are at the beginning after sorting. The proposed schemes were selected as the initial formal building population, and the scheme with the lowest fitness value was designated as the current top-level landmark blueprint. .

[0021] The master-led structural correction strategy in step S22 includes the following steps: Step S221: For each current scheme in the population Select all individuals from the population with fitness values ​​greater than [a certain value]. The proposed solutions constitute a set of superior solutions, and a solution is randomly selected from this set as a reference blueprint. ; Step S222: Utilize the top-level landmark blueprint in the current population. and the reference blueprint selected in step S221 Jointly guide the current plan The revised position update formula is as follows: , In the formula Indicates the location of the new scheme after structural modification. Indicates the location of the current building plan. Represents a random number between 0 and 1. This indicates that the top-level landmark blueprint represents the globally optimal solution. This refers to a random solution within the set of superior solutions, which is the reference blueprint.

[0022] The dynamic accuracy fine-tuning process strategy in step S23 includes the following steps: Step S231: Calculate the results as the number of iterations increases. Construction accuracy radius decreases exponentially , , In the formula This indicates the construction accuracy radius for the current iteration. This represents the initial radius scaling factor. Indicates the upper bound of the search space. Indicates the lower bound of the search space. Represents the precision convergence rate coefficient. Indicates the current iteration number. Indicates the maximum number of iterations; Step S232, Based on construction accuracy radius Apply isotropic random perturbations to the architectural design. , In the formula Indicates the location of the new design after fine polishing. Indicates the location of the current building plan. It represents a random vector between 0 and 1.

[0023] The reverse thinking design jump strategy in step S24 includes the following steps: Step S241: Generate a random number in each iteration and match the random number with the preset design jump probability. Compare the results to determine if the jump condition is met; Step S242: If the random number is less than The improved builder optimization algorithm calculates the current top-level landmark blueprint. Reverse design solutions within the search space , , In the formula This represents a reverse design scheme for a top-tier landmark blueprint. Indicates the lower bound of the search space. Indicates the upper bound of the search space. This represents the current blueprint for top-level landmarks; Step S243: Traverse the current building population and identify the solution with the highest fitness value as the discarded solution. ; Step S244: Calculate the reverse design scheme The fitness value, and compared with the discard scheme. Compare the fitness values, if The fitness value is less than The fitness value is then used. replace .

[0024] By implementing parameter tuning for a blood flow PID control system using the builder optimization algorithm and an improved builder optimization algorithm in Matlab, the following results are obtained: Figures 2-4 The comparison charts shown here compare the Builder Optimization Algorithm (BEM) and the improved BEM in terms of optimal fitness, step response, and PID parameter optimization process. The optimal fitness value for the BEM is 0.502758, while that for the improved BEM is 0.174157. The improved BEM finds a position closer to the global optimum in the objective function solution space, exhibiting higher optimization accuracy. Furthermore, the improved BEM continues to optimize the solution quality in the later stages of iteration, avoiding the premature stagnation phenomenon of the BEM, demonstrating stronger optimization capabilities. The rise time of the improved BEM is significantly shorter than that of the BEM; when switching blood flow from 2.5 L / min to 4.0 L / min, it reaches the target value of 4.0 L / min faster and has a shorter settling time. Figure 4As shown, the builder optimization algorithm exhibits premature convergence, while the improved builder optimization algorithm explores a wider solution space through sufficient search in the early and middle stages and can converge stably in the later stages. Therefore, it can be seen that the builder optimization algorithm has better performance, and the control parameters obtained by using the improved builder optimization algorithm have a better control effect on blood flow.

[0025] The following describes the specific process of parameter tuning for the blood flow PID control system of a medical extracorporeal circulation system using an improved builder optimization algorithm, obtaining the optimal proportional coefficient through optimization. Integral coefficient and differential coefficients This allows for rapid and stable control of blood flow.

[0026] I. System Modeling and Fitness Function Construction; Step 1: Construct the mathematical model and evaluation indicators for the PID control system of blood flow in the extracorporeal circulation monitoring system, specifically including: Step 11: Based on the blood flow control requirements of the extracorporeal circulation monitoring system, the extracorporeal circulation blood flow monitoring PID control system uses a PID control law to calculate the control quantity. , In the formula This indicates the control signal output to the blood flow regulation module. This indicates the real-time deviation between the set target blood flow and the monitored actual blood flow. This represents the proportionality coefficient. Represents the integral coefficient. Represents the differential coefficient; Step 12: The controlled object is the blood flow process of the extracorporeal circulation detection system. After system identification, its dynamic characteristics are described as a second-order transfer function with pure time delay. , In the formula The transfer function of the controlled object represents the dynamic response relationship between the blood pump speed and blood flow. The Laplace transform variables are represented by a system gain and a square term of natural frequency of 2.25, a damping coefficient of 2.4, and a pure time delay of 0.2. Step 13: Based on the characteristics of the extracorporeal circulation detection system, and following the principle of prioritizing safety while also considering speed, construct a fitness function for quantitatively evaluating the control effect of each set of PID parameters. , , In the formula This indicates that the simulation time is set to 15 seconds. Indicates systematic error. This indicates the overshoot penalty weight and takes a value of 1000. This indicates that the time penalty weight is adjusted and set to a value of 100. This indicates the actual overshoot of the system. This represents the target overshoot and has a value of 0.01. Indicates the actual adjustment time of the system. Indicates the target adjustment time and takes a value of 5 seconds.

[0027] II. Initialization; Step 2: Execute the dual-scheme pre-selection initialization strategy, which includes the following steps: Step 21: Set algorithm parameters: Set population size The value is set to 30, which sets the maximum number of iterations. Set the value to 100 to define the variable dimension. The value is 3. , , Set the lower bound of the search space A value of 0 sets the upper bound of the search space. The value is 30; Step 22: Generate a random initial building scheme And calculate the reverse building scheme. ; The improved builder optimization algorithm first in the search space 30 random individuals are generated internally. Then calculate the corresponding reverse individuals. , In the formula Represents a set of reverse architectural schemes; Step 23: Perform a hybrid screening based on fitness-based selection; The improved builder optimization algorithm will and The results are merged into a candidate set containing 60 individuals, and then the fitness function is used. The fitness value of each individual is calculated; the algorithm sorts all individuals by fitness value from smallest to largest, selects the top 30 individuals to form the formal initial building population, and marks the individual with the smallest fitness value as the current top-level landmark blueprint. .

[0028] III. Iterative optimization; Step 3: Enter the main loop and execute the master-led structural correction strategy, which includes the following steps: Step 31: For each individual in the population The algorithm traverses the population to find all fitness values ​​that are better than... Individuals constitute a set of superior solutions; Step 32: Randomly select one individual from the set of superior solutions as a reference blueprint. ; Step 33: Utilize the top-level landmark blueprint and reference blueprints Guide location update, , In the formula Indicates the new position after correction. Represents a random number between 0 and 1; Step 34, Calculation If its fitness value is better than The fitness value is then used. renew ; Step 4: Implement a dynamic precision fine-tuning process strategy, which includes the following steps: Step 41: Calculate the construction accuracy radius for the current iteration. , , In the formula With the number of iterations It exhibits exponential decay, ensuring that the search scope converges from the global to the local; Step 42, Based on construction accuracy radius To refine each individual piece meticulously. , In the formula This indicates the new position after fine-tuning. This represents a random vector of dimension 3; Step 43, Calculation If its fitness value is better than The fitness value is then used. renew ; Step 5: Implement a reverse thinking design jump strategy, which includes the following steps: Step 51: Generate a random number. If the random number is less than the intergenerational jump probability... ,in If the value is 0.1, then a jump operation is performed; Step 52: Calculate the reverse design scheme of the top-level landmark blueprint. , In the formula This represents the inverse solution generated based on the global optimal solution; Step 53: Identify the discarded scheme with the highest fitness value in the current population. ,like The fitness value is less than The fitness value is then used. replace .

[0029] IV. Termination and Output; Step 6: Determine the termination condition and output the result; The algorithm checks the current iteration count. Has the maximum number of iterations been reached? If not achieved, And return to step 3 to continue execution; If the target has been reached, the optimization process will terminate, and a top-level landmark blueprint will be output. The three corresponding values ​​are the optimal PID control parameters after tuning. =2.579477、 =2.409354、 =1.091096, and this parameter is applied to the PID control system for detecting blood flow in extracorporeal circulation to achieve the optimal blood flow regulation effect.

Claims

1. A blood flow control optimization method for an extracorporeal circulation detection system, characterized by, Comprising the following steps: Step S1, constructing a closed-loop feedback mechanism-based extracorporeal circulation blood flow PID control system, including a blood flow real-time monitoring module, a flow deviation calculation module, a blood flow PID controller module, a blood flow execution adjustment module, and an improved builder optimization algorithm parameter setting module; Step S2, constructing an improved builder optimization algorithm parameter setting module, the improvement of the improved builder optimization algorithm specifically including: Step S21, introducing a double scheme pre-selection initialization strategy, by executing random initial building scheme generation, reverse building scheme calculation, and mixed selection based on fitness value, introducing reverse solutions and selecting the best to reserve in the initialization stage, and expanding the initial search range to the complementary area; Step S22, introducing a master navigation structure correction strategy, by setting a difference guided position update formula based on the top landmark blueprint and the reference blueprint in the exploration stage, using the guiding effect of the global optimal solution and the random superior solution in the group to determine the search direction of structure correction; Step S23, introducing a dynamic precision fine-tuning process strategy, by setting a construction precision radius calculation formula that decays exponentially with construction progress and a local position update formula based on the construction precision radius in the development stage, controlling the disturbance range to nonlinearly shrink with the increase of iteration number, realizing the transition from global breadth search to local depth development; Step S24, introducing a reverse thinking design jump strategy, by setting a generation jump based on probability triggering, a reverse design scheme calculation of the top landmark blueprint, and a best replacement for the last individual in the population, introducing a reverse disturbance for the global optimal solution and replacing the worst individual, and using the complementary characteristics of the reverse solution to increase the diversity of the population; Step S3, using the improved builder optimization algorithm to set the blood flow PID control parameters in the extracorporeal circulation blood flow PID control system, and outputting the optimal control parameters through iterative optimization, the optimal control parameters including a proportional coefficient , an integral coefficient , and a differential coefficient ; Step S4, load the three optimal control parameters obtained by using the improved builder optimization algorithm to the blood flow PID controller module in the extracorporeal circulation blood flow PID control system, and the blood flow PID controller module performs PID control law operation based on the optimal control parameters to calculate the real-time deviation between the set target blood flow and the monitored actual blood flow to generate a control signal , and drive the blood flow adjustment module through the control signal to achieve fast response and high-precision steady-state tracking of the actual blood flow to the target flow.

2. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 1, wherein, In the extracorporeal circulation blood flow detection PID control system constructed in step S1, the real-time blood flow monitoring module is used to collect the instantaneous blood flow signal in the extracorporeal circulation tubing and transmit it as a feedback quantity to the flow deviation calculation module. The flow deviation calculation module receives the set target blood flow and calculates the real-time deviation between the target value and the feedback quantity. The data is then transmitted to the blood flow PID controller module. The improved builder optimization algorithm parameter tuning module serves as the upper-level optimization unit, calculating the fitness function of the system through iterative optimization. Minimize the optimal parameter vector [ , , The data is then loaded into the blood flow PID controller module, which uses the received optimal parameters and real-time deviations. The control quantity is calculated based on the PID control law. It then outputs the control quantity to the blood flow control module, which responds to the control quantity. By changing the pumping state to adjust the actual blood flow, precise tracking and stable control of the target flow can be achieved.

3. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 2, wherein, The control law adopted by the blood flow PID controller module is: ， In the formula represents a control output signal, represents a deviation between a set blood flow and an actual blood flow, represents a proportional coefficient, represents an integral coefficient, represents a differential coefficient.

4. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 1, wherein, The double scheme pre-selection initialization strategy in step S21 includes the following steps: Step S211, generating a random initial building scheme of size in the search space , ， wherein denotes a lower bound of the search space, denotes an upper bound of the search space, denotes a random vector between 0 and 1; Step S212, based on the random initial building scheme calculating the corresponding reverse building scheme , ， wherein denotes a reverse building scheme, denotes a search space lower bound, denotes a search space upper bound, denotes a random initial building scheme; Step S213, merging the random initial building scheme with the reverse building scheme to form a mixed candidate scheme set with a size of ​ Step S214, using fitness function Each scheme in the mixed candidate scheme set is evaluated to obtain a fitness value of each scheme, and all schemes are sorted in ascending order of fitness value. Step S215, select the scheme ranked in the front of the schemes as the official initial building population, and mark the scheme with the minimum fitness value as the current top landmark blueprint .

5. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 1, wherein, The master navigation structure correction strategy in step S22 includes the following steps: Step S221, for each current solution in the population , all solutions with fitness value better than are filtered out from the population to form a superior solution set, and a solution is randomly selected from the superior solution set as the reference blueprint ; Step S222, using the top landmarks blueprint in the current population and the reference blueprint selected in step S221 to jointly guide the amendment of the current solution The new position update formula is ， wherein represents the new solution position after structure modification, represents the current building solution position, represents a random number between 0 and 1, represents the top landmark blueprint, i.e., the global optimal solution, represents a reference blueprint, i.e., a random solution in the superior solution set.

6. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 1, wherein, The dynamic precision fine-tuning process strategy in step S23 includes the following steps: Step S231, calculating the iteration number Construction accuracy radius decaying exponentially , ， In the formula represents the construction accuracy radius of the current iteration, represents the initial radius proportionality coefficient, represents the upper bound of the search space, represents the lower bound of the search space, represents the accuracy convergence rate coefficient, represents the current iteration number, represents the maximum iteration number; Step S232, based on construction accuracy radius Performing isotropic random perturbation on the building scheme, ， wherein represents the new refined design position, represents the current building design position, represents a random vector between 0 and 1.

7. The blood flow control optimization method for extracorporeal circulation detection system according to claim 1, wherein, The reverse thinking design jump strategy in step S24 includes the following steps: Step S241, a random number is generated in each iteration, and the random number is compared with a preset design jump probability Comparison is made to determine whether the jump condition is met; Step S242, if the random number is less than , the improved builder optimization algorithm calculates the current top landmark blueprint Reverse design scheme within the search space , ， In the formula represents a reverse design scheme of a top landmark blueprint, represents a search space lower bound, represents a search space upper bound, represents a current top landmark blueprint; Step S243, traverse the current building population, and identify the scheme with the maximum fitness value as the abandoned scheme ; Step S244, calculate the fitness value of the reverse design scheme, and compare it with the fitness value of the discarded scheme If the fitness value of the discarded scheme is smaller than the fitness value of the reverse design scheme, replace the discarded scheme with the reverse design scheme ​​​​​ 8. The blood flow control optimization method for an extracorporeal circulation monitoring system according to claim 1, wherein, The improved builder optimization algorithm in step S3 uses a composite performance index as the fitness function , ， wherein denotes a fitness value, denotes a simulation time, denotes a system error, denotes an overshoot penalty weight, denotes a regulation time penalty weight, denotes an actual overshoot, denotes a target overshoot, denotes an actual regulation time, denotes a target regulation time.

Images