Multi-sensor pressure balance control method and system for bladder irrigation

Abstract

The application provides a multi-sensor pressure balance control method and system for bladder irrigation, relates to the technical field of pressure balance control, and comprises the following steps: calculating a pressure gradient by synchronously collecting pressure values of input and output ends of a perfusion pipeline, and outputting a flow regulation increment to drive a proportional regulating valve by using an adaptive fuzzy control mechanism based on gradient deviation and its change rate. Meanwhile, the residual volume of a cavity is dynamically estimated according to the cumulative flow difference between the input and the output, and a volume compensation coefficient is generated. Finally, the compensation coefficient and the flow regulation increment are weighted and fused to form a composite control quantity to cooperatively regulate the flow of the double pipeline and maintain the pressure stability in the cavity. The method realizes adaptive and accurate balance control of the cavity pressure in the bladder irrigation process, and improves the safety and stability of the operation.

Description

Technical Field

[0001] This invention relates to the field of pressure balance control technology, and in particular to a multi-sensor pressure balance control method and system for bladder irrigation. Background Technology

[0002] In urological surgeries, especially transurethral surgeries, bladder irrigation is a crucial procedure for maintaining a clear surgical view and a stable internal environment. Current techniques generally employ a single-point pressure feedback-based control method to stabilize intrabladder pressure. Specifically, the conventional approach involves placing a single pressure sensor directly in the irrigation output tubing or within the bladder cavity to monitor the intrabladder pressure. The control system compares this real-time monitored pressure value with a preset target pressure threshold. If a pressure deviation is detected, the system adjusts the peristaltic pump speed or the opening and closing of a valve in the irrigation input tubing to alter the infusion fluid flow rate, attempting to bring the intrabladder pressure back to the target range. The core logic of this method is single-loop negative feedback control, which is relatively simple in design and primarily relies on unidirectional adjustment of the input flow rate to respond to pressure changes.

[0003] Because it relies on only a single pressure monitoring point, the system cannot perceive the dynamic pressure relationship between the input and output ends of the irrigation circuit, and is slow to respond to upstream and downstream pressure imbalances caused by output tubing blockage, changes in patient position, or alterations in bladder compliance. This often leads to delayed control actions, resulting in significant fluctuations in intrabladder pressure. Sometimes the pressure is too high, increasing the risk of bladder perforation or fluid absorption, while sometimes it is too low, affecting surgical field clarity. A more prominent problem is that this method only focuses on instantaneous pressure regulation, lacking continuous estimation and safety management of the net retention of irrigation fluid within the surgical cavity. Even with reduced input flow, fluid continues to accumulate in the cavity when the output tubing is blocked, posing a potential risk of bladder overdistension and even rupture. A simple single-point pressure control mechanism cannot provide effective early warning and proactive intervention for this, making its safety and adaptability insufficient. Summary of the Invention

[0004] The present invention provides a multi-sensor pressure balance control method and system for bladder irrigation, which can solve the problems in the prior art.

[0005] A first aspect of the present invention provides a multi-sensor pressure balance control method for bladder irrigation, comprising: The pressure sensing units installed in the irrigation input pipeline and the irrigation output pipeline synchronously collect the input pressure value and the output pressure value, and calculate the pressure gradient value between the two. An adaptive fuzzy control mechanism based on pressure gradient value is constructed. The gradient deviation between the pressure gradient value and the target gradient value, as well as the time derivative of the gradient deviation value, are used as dual-input fuzzy variables. The dual-input fuzzy variables are mapped to the fuzzy domain through a membership function. Fuzzy inference is performed according to a preset fuzzy rule base, and the output flow adjustment increment is output after defuzzification processing. Based on the flow rate adjustment increment, a pulse width modulation signal is generated to drive the proportional regulating valve set in the irrigation input pipeline to adjust its valve opening, thereby synchronously controlling the input flow rate of the irrigation fluid. A dynamic estimation mechanism for surgical cavity volume is established. The amount of irrigation fluid retained in the surgical cavity is calculated based on the difference between the cumulative input flow rate of the irrigation input pipeline and the cumulative output flow rate of the irrigation output pipeline. The amount of retained fluid is compared with a preset volume safety boundary. When the amount of retained fluid approaches the volume safety boundary, a volume compensation coefficient is generated. The volume compensation coefficient and the flow rate adjustment increment are weighted and fused to obtain a composite control quantity. Based on the composite control quantity, the flow rate ratio of the irrigation input pipeline and the irrigation output pipeline is adjusted in a coordinated manner to maintain the intraoperative pressure within a preset steady-state range.

[0006] Constructing an adaptive fuzzy control mechanism based on pressure gradient values ​​includes: The pressure gradient value is sampled by a sliding window to obtain a pressure gradient sequence. The pressure gradient mean and pressure gradient variance are calculated based on the pressure gradient sequence. The pressure gradient variance is used to determine whether the current pressure fluctuation state belongs to a steady state mode or a disturbance mode. In the steady-state mode, the dual-input fuzzy variables are quantized using a first universe of discourse range, and in the perturbation mode, the dual-input fuzzy variables are quantized using a second universe of discourse range, wherein the second universe of discourse range is larger than the first universe of discourse range. The input quantization factor is dynamically calculated based on the ratio of the pressure gradient variance to the preset variance benchmark value. The gradient deviation value and the corresponding gradient deviation change rate are then multiplied by the input quantization factor and mapped to the corresponding domain range. A Gaussian membership function set is constructed for the quantized dual-input fuzzy variables. Each language variable level corresponds to a Gaussian membership function, and the Gaussian membership functions of adjacent language variable levels maintain a preset overlap. Based on the fuzzy rule base, inference operations are performed to obtain the fuzzy output. The corresponding output scaling factor is selected according to the pressure fluctuation state to scale the defuzzification result and generate the flow regulation increment.

[0007] A Gaussian membership function set is constructed for the quantized dual-input fuzzy variables. Each linguistic variable level corresponds to one Gaussian membership function. The Gaussian membership functions of adjacent linguistic variable levels maintain a preset overlap, including: The number of language variable levels is determined based on the domain range of the dual-input fuzzy variables. The domain range is then uniformly divided into multiple sub-intervals corresponding to the number of language variable levels. The center point of each sub-interval is used as the peak center of the Gaussian membership function of the corresponding language variable level. The width parameter of the Gaussian membership function is calculated based on the preset overlap between adjacent linguistic variable levels, and the width parameter determines the decay rate of the Gaussian membership function. A first Gaussian membership function group is constructed for the gradient deviation value. Each Gaussian membership function in the first Gaussian membership function group shares the same width parameter and the peak centers are arranged sequentially according to the center points of the sub-intervals. A second Gaussian membership function set is constructed for the gradient deviation change rate, and the construction method of the second Gaussian membership function set is the same as that of the first Gaussian membership function set.

[0008] Based on the flow rate adjustment increment, a pulse width modulation signal is generated to drive a proportional regulating valve installed in the irrigation input pipeline to adjust its valve opening, including: The flow rate adjustment increment is rate-limited. The flow rate difference between the current flow rate adjustment increment and the previous cycle flow rate adjustment increment is calculated. When the flow rate difference exceeds the preset increment change limit, the flow rate adjustment increment is corrected to the limit boundary to generate a rate-limited flow rate adjustment increment. The rate-limited flow regulation increment is reverse-compensated based on the flow characteristic curve of the proportional control valve to eliminate the influence of the valve's inherent nonlinear characteristics on the flow control accuracy. The target duty cycle is calculated based on the compensated flow regulation increment, and the pulse width modulation signal is generated by combining it with the preset modulation frequency. The duty cycle of the pulse width modulation signal is consistent with the target duty cycle. A jitter component is superimposed on the pulse width modulation signal, the amplitude of which is less than the duty cycle resolution; the pulse width modulation signal after superimposing the jitter component is amplified and then output to the proportional control valve to drive the proportional control valve to adjust its valve opening.

[0009] When the retention amount approaches the volume safety boundary, a volume compensation coefficient is generated, including: Linear fitting is performed on the continuously collected retention data to obtain the slope of retention change. Based on the current retention and the slope of retention change, the predicted retention after a preset time window is predicted. The difference between the preset volume safety boundary and the predicted retention amount is used as the prediction margin value; The current response level is determined based on the range in which the predicted margin value falls, and the response level includes normal level, attention level and emergency level. Different compensation coefficient generation strategies are configured for different response levels. The regular level corresponds to a unit compensation coefficient, the attention level corresponds to a gradual compensation coefficient based on linear interpolation of the prediction margin value, and the emergency level corresponds to an enhanced compensation coefficient based on exponentially increasing prediction margin value. The compensation coefficient generation strategy corresponding to the current response level is applied to the prediction margin value, and the volume compensation coefficient is output.

[0010] Determining the current response level based on the range of the predicted margin value includes: Set a first margin threshold and a second margin threshold, where the first margin threshold is greater than the second margin threshold, and the first margin threshold and the second margin threshold divide the margin space into three continuous intervals; When the prediction margin value is greater than the first margin threshold, the current response level is determined to be the normal level, and the system is in a safe operating state. When the predicted margin value is less than or equal to the first margin threshold and greater than the second margin threshold, the current response level is determined to be the level of concern, and a gradual pressure control strategy is initiated. When the predicted margin value is less than or equal to the second margin threshold, the current response level is determined to be the emergency level, and the forced decompression protection strategy is activated. A response level migration hysteresis mechanism is established. When the response level migrates from high to low, the first margin threshold and the second margin threshold are used as the judgment boundary. When the response level recovers from low to high, the threshold after adding a preset hysteresis amount is used as the judgment boundary.

[0011] Based on the aforementioned composite control quantity, the flow ratio of the irrigation input pipeline and the irrigation output pipeline is coordinated and adjusted, including: The composite control quantity is decomposed into a pressure maintenance component and a volume balance component. The pressure maintenance component is used to adjust the pressure level in the surgical cavity, and the volume balance component is used to adjust the total amount of irrigation fluid in the surgical cavity. The main regulating pipeline is determined based on the deviation direction between the current intraoperative pressure value and the center value of the preset steady-state range. When the intraoperative pressure value is higher than the center value, the irrigation output pipeline is determined as the main regulating pipeline, and when the intraoperative pressure value is lower than the center value, the irrigation input pipeline is determined as the main regulating pipeline. Calculate the flow ratio between the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation, and apply pipeline flow constraints to the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation respectively to ensure that the flow regulation of each pipeline does not exceed the maximum flow capacity and minimum flow capacity of the corresponding pipeline. The constrained flow rate adjustment is simultaneously sent to the actuators of both the irrigation input pipeline and the irrigation output pipeline to achieve coordinated regulation of the two pipelines.

[0012] A second aspect of the present invention provides a multi-sensor pressure balancing control system for bladder irrigation, comprising: The pressure gradient unit is used to synchronously acquire the input pressure value and the output pressure value through the pressure sensing unit installed in the irrigation input pipeline and the irrigation output pipeline, and to calculate the pressure gradient value between the two. The fuzzy control unit is used to construct an adaptive fuzzy control mechanism based on the pressure gradient value. The gradient deviation between the pressure gradient value and the target gradient value, as well as the time derivative of the gradient deviation value, are used as dual-input fuzzy variables. The dual-input fuzzy variables are mapped to the fuzzy domain through a membership function. Fuzzy inference is performed according to a preset fuzzy rule base, and the flow rate adjustment increment is output after defuzzification processing. The pulse width modulation unit is used to generate a pulse width modulation signal according to the flow rate adjustment increment, drive the proportional regulating valve set in the irrigation input pipeline to adjust its valve opening, and synchronously regulate the input flow rate of the irrigation fluid. The volume estimation unit is used to establish a dynamic estimation mechanism for the surgical cavity volume. It calculates the amount of irrigation fluid retained in the surgical cavity based on the difference between the cumulative input flow rate of the irrigation input pipeline and the cumulative output flow rate of the irrigation output pipeline. The retained amount is compared with a preset volume safety boundary. When the retained amount approaches the volume safety boundary, a volume compensation coefficient is generated. The composite control unit is used to weight and fuse the volume compensation coefficient and the flow rate adjustment increment to obtain a composite control quantity, and based on the composite control quantity, coordinately adjust the flow rate ratio of the irrigation input pipeline and the irrigation output pipeline to maintain the intraoperative cavity pressure within a preset steady-state range.

[0013] A third aspect of the present invention provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0015] The beneficial effects of this application are as follows: This method achieves precise balanced control of intraoperative cavity pressure during bladder irrigation. By synchronously acquiring pressure values ​​at both ends of the irrigation input and output tubing and calculating the pressure gradient in real time, it can directly reflect the dynamic change trend of intraoperative cavity pressure, providing accurate feedback signals for closed-loop control. The constructed adaptive fuzzy control mechanism uses pressure gradient deviation and its rate of change as dual input variables, effectively integrating the static error and dynamic trend of pressure state. Through fuzzy inference and defuzzification processing of the output flow regulation increment, the control decision possesses robustness and adaptability to cope with nonlinear and time-varying system characteristics, avoiding the shortcomings of traditional PID control in terms of difficult parameter tuning and easy overshoot oscillation.

[0016] A pulse-width modulation (PWM) signal was introduced to drive a proportional control valve, enabling continuous, smooth, and high-resolution regulation of the perfusion fluid input flow rate. This overcomes the flow rate abrupt changes and pressure shocks associated with traditional on / off valve control, significantly improving the stability and response speed of pressure regulation. A dynamic estimation mechanism for the surgical cavity volume was established. By calculating the amount of perfusion fluid retained in the surgical cavity in real time through cumulative flow differences, it can proactively predict potential risks from pressure changes and generate a volume compensation coefficient when the retained volume approaches a safe boundary. This mechanism combines pressure feedback control with volume feedforward compensation, forming a composite control strategy. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of a multi-sensor pressure balance control method for bladder irrigation according to an embodiment of the present invention; Figure 2 This is a flowchart of the pressure prediction and compensation control strategy for the bladder irrigation system according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0020] Figure 1 This is a flowchart illustrating a multi-sensor pressure balance control method for bladder irrigation according to an embodiment of the present invention. Figure 1 As shown, the method includes: The pressure sensing units installed in the irrigation input pipeline and the irrigation output pipeline synchronously collect the input pressure value and the output pressure value, and calculate the pressure gradient value between the two. An adaptive fuzzy control mechanism based on pressure gradient value is constructed. The gradient deviation between the pressure gradient value and the target gradient value, as well as the time derivative of the gradient deviation value, are used as dual-input fuzzy variables. The dual-input fuzzy variables are mapped to the fuzzy domain through a membership function. Fuzzy inference is performed according to a preset fuzzy rule base, and the output flow adjustment increment is output after defuzzification processing. Based on the flow rate adjustment increment, a pulse width modulation signal is generated to drive the proportional regulating valve set in the irrigation input pipeline to adjust its valve opening, thereby synchronously controlling the input flow rate of the irrigation fluid. A dynamic estimation mechanism for surgical cavity volume is established. The amount of irrigation fluid retained in the surgical cavity is calculated based on the difference between the cumulative input flow rate of the irrigation input pipeline and the cumulative output flow rate of the irrigation output pipeline. The amount of retained fluid is compared with a preset volume safety boundary. When the amount of retained fluid approaches the volume safety boundary, a volume compensation coefficient is generated. The volume compensation coefficient and the flow rate adjustment increment are weighted and fused to obtain a composite control quantity. Based on the composite control quantity, the flow rate ratio of the irrigation input pipeline and the irrigation output pipeline is adjusted in a coordinated manner to maintain the intraoperative pressure within a preset steady-state range.

[0021] The ultimate goal of this invention is to precisely control intrabladder pressure using an adaptive fuzzy control algorithm. Figure 1 In its implementation, the system not only synchronously collects pressure values ​​and calculates pressure gradients through pressure sensing units installed in the irrigation input and output lines, but also monitors the amount of irrigation fluid retained in the bladder by combining a dynamic estimation mechanism of the surgical cavity volume. Considering the individual differences in bladder capacity among patients, the system uses real-time intrabladder pressure as the primary control parameter, while using the amount of irrigation fluid retained as an auxiliary indicator. Through the dynamic ratio of the pressure maintenance component and the volume balance component, precise control of intrabladder pressure is achieved. This dual feedback mechanism ensures the accuracy of pressure control while providing over-limit warning protection, better adapting to the individualized needs of different patients.

[0022] In one alternative implementation, constructing an adaptive fuzzy control mechanism based on pressure gradient values ​​includes: The pressure gradient value is sampled by a sliding window to obtain a pressure gradient sequence. The pressure gradient mean and pressure gradient variance are calculated based on the pressure gradient sequence. The pressure gradient variance is used to determine whether the current pressure fluctuation state belongs to a steady state mode or a disturbance mode. In the steady-state mode, the dual-input fuzzy variables are quantized using a first universe of discourse range, and in the perturbation mode, the dual-input fuzzy variables are quantized using a second universe of discourse range, wherein the second universe of discourse range is larger than the first universe of discourse range. The input quantization factor is dynamically calculated based on the ratio of the pressure gradient variance to the preset variance benchmark value. The gradient deviation value and the corresponding gradient deviation change rate are then multiplied by the input quantization factor and mapped to the corresponding domain range. A Gaussian membership function set is constructed for the quantized dual-input fuzzy variables. Each language variable level corresponds to a Gaussian membership function, and the Gaussian membership functions of adjacent language variable levels maintain a preset overlap. Based on the fuzzy rule base, inference operations are performed to obtain the fuzzy output. The corresponding output scaling factor is selected according to the pressure fluctuation state to scale the defuzzification result and generate the flow regulation increment.

[0023] In establishing an adaptive fuzzy control mechanism, continuous monitoring and processing of pressure gradient values ​​is fundamental to achieving precise control. Pressure gradient values ​​are sampled using a sliding window of length N, typically ranging from 5 to 15 sampling points, with the specific value determined based on the system's response characteristics. During sampling, pressure gradient values ​​are acquired sequentially according to a time series, forming a pressure gradient sequence g_1, g_2, ..., g_N, where g_i represents the pressure gradient value at the i-th sampling time. The mean pressure gradient is then calculated based on this sequence. As a basis for assessing the current trend of pressure change, the pressure gradient variance is also calculated. This reflects the severity of pressure fluctuations. When the variance value is below the set steady-state threshold... When the current pressure fluctuation is determined to be in steady state mode, the system operates smoothly and the pressure changes relatively regularly. When the variance value exceeds the threshold, it is determined to be in disturbance mode, indicating that there is external interference or sudden load change that causes a large fluctuation in pressure.

[0024] Differentiated universe of discourse (UD) settings are employed for different pressure fluctuation states to ensure the fuzzy controller maintains good performance under various operating conditions. In steady-state mode, the UD range of the dual-input fuzzy variables is set to the first UD range, which is relatively narrow. For example, the UD range of the gradient deviation value is set to [-2, 2], and the UD range of the gradient deviation rate of change is set to [-1.5, 1.5]. A narrower UD range improves the control resolution in steady state, enabling the controller to produce a fine response to small deviation changes. In disturbance mode, the UD range is expanded to the second UD range, typically by a factor of 1.5 to 2. For example, the UD range of the gradient deviation value is expanded to [-4, 4], and the UD range of the gradient deviation rate of change is expanded to [-3, 3]. A wider UD range can cover larger input changes under disturbance conditions, preventing input saturation that could lead to control failure.

[0025] To achieve an accurate mapping between the universe of discourse and the actual input value, an input quantization factor is introduced for dynamic adjustment. The variance of the pressure gradient is calculated and compared with a preset variance benchmark value. ratio This ratio directly reflects the degree of deviation of the current fluctuation from normal operating conditions. Based on this ratio, the input quantization factor is calculated using a piecewise function or a continuous function. The typical calculation method is as follows ,in The baseline quantization factor is α, which is an adjustment coefficient typically ranging from 0.2 to 0.5. The actual measured gradient deviation value e is compared with the gradient deviation rate of change. Multiply by the quantization factor to obtain the quantized value. and The quantification results are then mapped to the domain of discourse corresponding to the current pressure fluctuation state, thus completing the standardization of the input variables.

[0026] In the fuzzification stage, a Gaussian membership function set is constructed for the quantized dual-input fuzzy variables to achieve precise fuzzy representation. The universe of discourse for each input variable is divided into seven linguistic variable levels: negative large, negative medium, negative small, zero, positive small, positive medium, and positive large. Each level corresponds to a Gaussian membership function. The mathematical expression of the Gaussian membership function is as follows: ; Where c is the center position of the membership function. The standard deviation parameter of the Gaussian function determines the width of the membership function. Gaussian membership functions for adjacent linguistic variable levels maintain an overlap of 0.3 to 0.5. This overlap design ensures that the fuzzy controller can simultaneously activate multiple fuzzy rules when the input value is at the boundary between two linguistic variables, achieving a smooth transition and avoiding abrupt changes in the control output. Compared to triangular or trapezoidal membership functions, Gaussian membership functions are continuously differentiable, providing a smoother fuzzy inference process.

[0027] A complete fuzzy rule base is established to describe the control logic relationship between inputs and outputs. The rule base is constructed in an "if-then" format. A typical rule form is "If the gradient deviation value is positive and the gradient deviation change rate is positive, then the flow regulation increment is positive," indicating that when the pressure gradient deviation is large and continues to increase, a larger positive flow regulation needs to be applied. The rule base contains 49 rules, covering 7×7 input combinations. When performing fuzzy inference operations, the membership degree of each linguistic variable is first calculated based on the current input value using a membership function. Then, fuzzy rules that meet the conditions are activated, and rule inference is performed using the minimum value method or the product method. The inference results of each rule are then aggregated to obtain the fuzzy output.

[0028] To obtain precise control commands, the fuzzy output is defuzzified. A common defuzzification method is the centroid method, which calculates the x-coordinate of the weighted centroid of all output membership functions as the clear output value. After obtaining the defuzzification result, an appropriate output scaling factor is selected for scaling adjustment based on the current pressure fluctuation state. A smaller output scaling factor is used in steady-state mode. The typical value range is 0.5 to 0.8. This setting makes the control action more conservative in steady state, reducing unnecessary flow adjustments and improving system stability. A larger output scaling factor is used in disturbance mode. The value ranges from 1.2 to 1.8. Increasing the scaling factor can produce a stronger control effect and quickly suppress the influence of disturbances.

[0029] In one optional implementation, a Gaussian membership function set is constructed for the quantized dual-input fuzzy variables, with each linguistic variable level corresponding to one Gaussian membership function. Maintaining a preset overlap between the Gaussian membership functions of adjacent linguistic variable levels includes: The number of language variable levels is determined based on the domain range of the dual-input fuzzy variables. The domain range is then uniformly divided into multiple sub-intervals corresponding to the number of language variable levels. The center point of each sub-interval is used as the peak center of the Gaussian membership function of the corresponding language variable level. The width parameter of the Gaussian membership function is calculated based on the preset overlap between adjacent linguistic variable levels, and the width parameter determines the decay rate of the Gaussian membership function. A first Gaussian membership function group is constructed for the gradient deviation value. Each Gaussian membership function in the first Gaussian membership function group shares the same width parameter and the peak centers are arranged sequentially according to the center points of the sub-intervals. A second Gaussian membership function set is constructed for the gradient deviation change rate, and the construction method of the second Gaussian membership function set is the same as that of the first Gaussian membership function set.

[0030] After completing the universe-of-discourse quantization of the dual-input fuzzy variables, a corresponding Gaussian membership function needs to be constructed for each linguistic variable level. First, the data distribution characteristics of the gradient deviation value and the gradient deviation rate of change are analyzed to determine the boundary values ​​of their respective universes of discourse. For the gradient deviation value, its universe of discourse range is assumed to be [-3.0, 3.0], and for the gradient deviation rate of change, its universe of discourse range is set to [-2.5, 2.5]. Considering the balance between the accuracy requirements of fuzzy control and computational complexity, the number of linguistic variable levels is determined to be 7, corresponding to seven semantic descriptions: negative large, negative medium, negative small, zero, positive small, positive medium, and positive large.

[0031] For the gradient bias value's universe of discourse [-3.0, 3.0], this interval is uniformly divided into 7 sub-intervals. The entire universe of discourse spans 6.0 units. After dividing it into 7 levels, the distance between the center points of adjacent levels is 6.0 divided by 6, which equals 1.0 units. The center points of these 7 sub-intervals are -3.0, -2.0, -1.0, 0.0, 1.0, 2.0, and 3.0, respectively. These center points will serve as the peak centers of each Gaussian membership function. At the peak center position, the membership degree of the corresponding linguistic variable level reaches its maximum value of 1.0, indicating that the input value completely belongs to that linguistic variable level.

[0032] The width parameter of the Gaussian membership function directly affects the degree of function broadening and decay rate. To ensure an appropriate transition region between adjacent linguistic variable levels and avoid membership functions that are too sharp or too flat, a preset overlap is set to 0.5. This overlap means that at the midpoint between two adjacent peak centers, the membership values ​​of the Gaussian membership functions on both sides reach 0.5. According to the mathematical properties of the Gaussian function, when the input value deviates from the peak center by a distance equal to 0.8326 times the width parameter, the membership value decays to 0.5. Since the distance between adjacent peak centers is 1.0, and the distance between the midpoint of two centers and any center is 0.5 units, the calculated width parameter is approximately 0.6 units. This parameter value ensures that the membership values ​​of adjacent membership functions at the midpoint are exactly 0.5, achieving the expected overlap effect.

[0033] When constructing the first Gaussian membership function set, seven Gaussian membership functions were established for each of the seven linguistic variable levels of gradient deviation. These functions all used a uniform width parameter of 0.6, with peak centers set to -3.0, -2.0, -1.0, 0.0, 1.0, 2.0, and 3.0, respectively. For the membership function of the negative large level, its peak center is located at -3.0. When the input gradient deviation is -3.0, the membership is 1.0. As the input value moves towards zero, the membership gradually decreases according to the Gaussian decay law. When the input value reaches -2.5, this value is exactly at the midpoint between the peak centers of the negative large and negative medium levels. At this point, the membership of the negative large level decays to 0.5, while the membership of the negative medium level is also 0.5, and their sum is 1.0, satisfying the completeness requirement of fuzzy partitioning. The same construction logic was applied sequentially to the remaining six linguistic variable levels: negative medium, negative small, zero, positive small, positive medium, and positive large.

[0034] In practical applications, when the gradient bias of the input is -1.8, this value lies between the negative medium and negative small levels. Calculated using Gaussian membership functions, this input value has a membership degree of approximately 0.73 for the negative medium level, approximately 0.24 for the negative small level, and close to zero for other levels. This gradual membership assignment avoids the boundary jump problem of hard classification, making the fuzzy inference process smoother and more continuous. The functions in the first Gaussian membership function set decay naturally at the boundaries of the universe of discourse, ensuring that boundary input values ​​still receive reasonable membership evaluations.

[0035] When constructing the second Gaussian membership function group for the gradient deviation rate of change, the construction method is completely consistent with that of the first group. Based on the universe of discourse range of the gradient deviation rate of change [-2.5, 2.5], the entire span is 5.0 units, also divided into 7 linguistic variable levels. The distance between adjacent peak centers is approximately 0.833 units (5.0 divided by 6). The 7 peak centers are -2.5, -1.667, -0.833, 0.0, 0.833, 1.667, and 2.5, respectively. Keeping the preset overlap at 0.5, the width parameter is recalculated based on the new center point distance, resulting in a width parameter of approximately 0.5 units. The 7 functions in the second Gaussian membership function group share this width parameter, and the peak centers are arranged sequentially according to the above calculation results, forming a complete membership function system covering the entire universe.

[0036] After constructing the two sets of Gaussian membership functions, they correspond to the two input dimensions of the fuzzy controller. When the actual collected gradient deviation value and gradient deviation rate of change are input, fuzzification processing is performed using the first and second Gaussian membership function sets respectively, converting the precise values ​​into membership vectors for each linguistic variable level. This membership design method based on Gaussian functions has good mathematical properties; the functions are smooth and continuous with derivatives, facilitating subsequent adaptive adjustment and optimization calculations. The appropriate overlap between adjacent functions ensures the continuity and robustness of fuzzy inference, avoiding drastic fluctuations in control output when the input value is near the linguistic variable boundary, providing a stable and reliable input fuzzification foundation for the entire fuzzy control system.

[0037] In one optional implementation, a pulse width modulation signal is generated based on the flow rate adjustment increment to drive a proportional regulating valve disposed in the irrigation input pipeline to adjust its valve opening, including: The flow rate adjustment increment is rate-limited. The flow rate difference between the current flow rate adjustment increment and the previous cycle flow rate adjustment increment is calculated. When the flow rate difference exceeds the preset increment change limit, the flow rate adjustment increment is corrected to the limit boundary to generate a rate-limited flow rate adjustment increment. The rate-limited flow regulation increment is reverse-compensated based on the flow characteristic curve of the proportional control valve to eliminate the influence of the valve's inherent nonlinear characteristics on the flow control accuracy. The target duty cycle is calculated based on the compensated flow regulation increment, and the pulse width modulation signal is generated by combining it with the preset modulation frequency. The duty cycle of the pulse width modulation signal is consistent with the target duty cycle. A jitter component is superimposed on the pulse width modulation signal, and the amplitude of the jitter component is smaller than the duty cycle resolution. The pulse width modulation signal with superimposed jitter components is amplified and then output to the proportional control valve to drive the proportional control valve to adjust its valve opening.

[0038] In practical irrigation control systems, the generation process of pulse width modulation (PWM) signals requires comprehensive consideration of multiple factors, including control smoothness, valve characteristic compensation, and execution accuracy. To prevent sudden changes in flow regulation commands from impacting the system, a rate-limiting process is first applied to the calculated flow regulation increment. Specifically, within each control cycle, the flow difference between the current cycle's flow regulation increment and the previous cycle's flow regulation increment is calculated. This difference reflects the rate of change of the control command. When the absolute value of the flow difference exceeds a preset increment change limit, the control command is considered to be changing too rapidly. In this case, the current flow regulation increment is corrected to the limit boundary value. That is, if the difference is positive and exceeds the limit, the current increment is set to the previous cycle's increment plus the positive limit value; if the difference is negative and exceeds the limit, it is set to the previous cycle's increment minus the negative limit value, thus generating a rate-limited flow regulation increment. This limit value is typically set to 2% to 5% of the rated flow rate to ensure that the flow change slope remains within a controllable range, avoiding mechanical shocks and fluid oscillations in the proportional control valve caused by sudden changes in commands.

[0039] The rate-limited flow regulation increment still requires further processing to address the inherent nonlinear characteristics of the proportional control valve. In practical applications, due to the valve core structure, sealing method, and hydrodynamic effects, the relationship between flow rate and valve opening in proportional control valves is not ideally linear. A flow characteristic curve for the proportional control valve is obtained through pre-calibration, describing the actual flow value corresponding to different valve openings. During control, the flow characteristic curve is consulted in reverse based on the rate-limited flow regulation increment to find the actual valve opening adjustment required to achieve the target flow change. This process essentially performs reverse compensation for the flow command, eliminating the influence of the valve's inherent nonlinear characteristics on flow control accuracy. For example, when the valve is in a small opening region, the flow gain is low, requiring a larger valve opening change for the same flow regulation increment; while in a large opening region, the flow gain is high, requiring a smaller opening change to achieve the same flow regulation effect.

[0040] After nonlinear compensation, the compensated flow regulation increment is mapped to the target duty cycle of the proportional control valve. This mapping relationship is based on the linear correspondence between valve opening and drive current. Precise valve opening adjustment is achieved by controlling the average drive current through the duty cycle. Specifically, the compensated flow regulation increment is normalized according to the valve's full-scale opening to obtain a relative opening value between 0 and 1. This relative opening value is directly used as the target duty cycle. A pulse width modulation (PWM) signal is generated by combining this with a preset modulation frequency, typically selected between 200Hz and 1kHz, to ensure control response speed while avoiding resonance in the valve's mechanical components. The generated PWM signal contains two phases within one modulation cycle: a high level and a low level. The ratio of the high-level duration to the total cycle time is the actual duty cycle, which strictly matches the target duty cycle calculated above.

[0041] To further improve the precision of flow control, especially under conditions of limited duty cycle resolution, a small jitter component is superimposed on the pulse width modulation (PWM) signal. This jitter component is a high-frequency disturbance signal with controlled amplitude, set to be less than the minimum resolution unit of the duty cycle, typically 0.3 to 0.7 times the duty cycle resolution. By introducing this controlled jitter, control precision exceeding hardware resolution limitations can be achieved statistically. The principle lies in the low-pass filtering effect of the valve's mechanical system on high-frequency disturbances, ensuring that the valve's actual response is the time-averaged effect of the PWM signal including the jitter. The frequency of the jitter component is usually set to 3 to 5 times the modulation frequency to ensure effective averaging over several modulation cycles. The jitter signal can be in the form of a pseudo-random sequence or a triangular wave, achieving spectral expansion while maintaining overall energy control.

[0042] The pulse width modulation (PWM) signal, after being superimposed with jitter components, is fed into a drive amplifier circuit. This circuit typically employs an H-bridge or push-pull topology, providing sufficient drive current to overcome the inductive effect and back electromotive force of the proportional control valve coil. The output current range of the drive amplifier circuit is generally 0 to 200 mA, and the output voltage is adapted to the rated operating voltage of the valve coil, commonly 12V or 24V DC. The PWM signal, after power amplification, is directly applied to the electromagnetic coil of the proportional control valve. The average value of the coil current is determined by the duty cycle of the PWM signal, thereby generating a corresponding electromagnetic force to drive the valve core to move, ultimately achieving precise adjustment of the valve opening. The entire execution chain, from flow regulation increment to valve opening adjustment, ensures control smoothness through rate limiting, improves control accuracy through nonlinear compensation, and overcomes resolution limitations through jitter injection, thus achieving high-precision, low-oscillation percolation flow control.

[0043] In one optional implementation, generating a volume compensation coefficient when the retention amount approaches the volume safety boundary includes: Linear fitting is performed on the continuously collected retention data to obtain the slope of retention change. Based on the current retention and the slope of retention change, the predicted retention after a preset time window is predicted. The difference between the preset volume safety boundary and the predicted retention amount is used as the prediction margin value; The current response level is determined based on the range in which the predicted margin value falls, and the response level includes normal level, attention level and emergency level. Different compensation coefficient generation strategies are configured for different response levels. The regular level corresponds to a unit compensation coefficient, the attention level corresponds to a gradual compensation coefficient based on linear interpolation of the prediction margin value, and the emergency level corresponds to an enhanced compensation coefficient based on exponentially increasing prediction margin value. The compensation coefficient generation strategy corresponding to the current response level is applied to the prediction margin value, and the volume compensation coefficient is output.

[0044] like Figure 2 As shown, the method includes: During the operation of an electrolyzer, real-time monitoring and dynamic compensation of electrolyte retention are crucial for maintaining system stability. By continuously collecting and predicting retention data, preventative compensation can be implemented before the retention approaches the safe volume boundary, thus avoiding abnormal electrolyzer operation caused by excessive retention.

[0045] A data acquisition module continuously monitors the current retention rate at a preset sampling frequency, typically set to once per second to once per minute, with the specific value determined based on the electrolyzer volume and electrolyte circulation rate. For medium-sized electrolyzers with volumes ranging from 500 liters to 2000 liters, the sampling interval is preferably every 5 to 15 seconds. The continuously collected retention data are arranged in chronological order to form a retention time series. The most recent preset number of data points is selected to construct an analysis window, typically ranging from 10 to 30.

[0046] The least squares method is applied to linearly fit the retention data points within the analysis window to obtain the slope of the retention change. The linear fitting process establishes a linear relationship between retention and time; the slope of the fitted line reflects the rate of change of retention over time. A positive slope indicates an upward trend in retention; a negative slope indicates a downward trend; the absolute value of the slope represents the rate of change. Using the current retention as the prediction starting point, and combining it with the slope of retention change, the predicted retention after a preset time window is calculated through linear extrapolation. This preset time window is set according to the electrolytic cell process characteristics, typically ranging from 30 to 300 seconds. For faster-responding electrolytic systems, the time window can be shortened to 10 to 60 seconds.

[0047] After obtaining the predicted retention amount, the difference between the preset volumetric safety boundary value and the predicted retention amount is calculated to obtain the prediction margin value. The prediction margin value reflects the distance between the predicted retention amount and the safety boundary; the smaller the value, the closer the system is to a critical safety state. The volumetric safety boundary is usually set to 80% to 95% of the effective volume of the electrolyzer, with the specific value determined according to the electrolyzer design specifications and process safety requirements. For example, for an electrolyzer with a rated volume of 1000 liters, if the volumetric safety boundary is set to 900 liters, when the predicted retention amount is 850 liters, the prediction margin value is calculated to be 50 liters.

[0048] Based on the predicted margin value's numerical range, the system's operating status is divided into three response levels. This level division is achieved by setting two thresholds: the first threshold separates the normal level from the concern level, and the second threshold separates the concern level from the emergency level. When the predicted margin value is greater than the first threshold, it is classified as normal, indicating that the system is operating within a safe range; when the predicted margin value is between the first and second thresholds, it is classified as concern, indicating that the system is approaching the safety boundary; when the predicted margin value is less than the second threshold, it is classified as emergency, indicating that the system is about to reach or has already reached a critical state. The first threshold is typically set to 10% to 20% of the volumetric safety boundary, and the second threshold is typically set to 3% to 8% of the volumetric safety boundary. Taking an electrolyzer with a volumetric safety boundary of 900 liters as an example, the first threshold can be set to 135 liters, and the second threshold can be set to 45 liters.

[0049] For three different response levels, differentiated compensation coefficient generation strategies are configured. For the normal level, the system operates smoothly and is far from the safety boundary, requiring no additional volumetric compensation adjustments; therefore, a unit compensation coefficient is configured, i.e., the compensation coefficient is set to 1.0, keeping the basic compensation scheme unchanged. For the concern level, the system begins to approach the safety boundary but still has sufficient response time. A linear interpolation strategy based on the prediction margin value is used to generate a gradual compensation coefficient. This strategy normalizes the prediction margin value between a first threshold and a second threshold, with the mapping result varying from 0 to 1, and then converts this mapping result into a compensation coefficient. The compensation coefficient increases linearly as the prediction margin value decreases, typically gradually increasing from 1.0 to a preset upper limit between 1.5 and 2.0. For the emergency level, the system approaches or reaches a critical safety state requiring rapid enhanced compensation. An exponentially increasing strategy based on the prediction margin value is used to generate enhanced compensation coefficients. This strategy uses the ratio of the prediction margin value to the second threshold as the independent variable of an exponential function, calculating the compensation coefficient through an exponential decay function. As the prediction margin approaches zero, the compensation coefficient grows exponentially, with an upper limit set at 2.5 to 5.0 to ensure the system has sufficient compensation response capability under extreme conditions.

[0050] After determining the current response level, the corresponding compensation coefficient generation strategy is invoked. The predicted margin value is used as the input parameter of the strategy function, and after corresponding mathematical operations, the final volume compensation coefficient is output. This volume compensation coefficient is then applied to the electrolyte replenishment control logic. By adjusting the flow rate of the replenishment pump or the opening of the replenishment valve, the dynamic balance adjustment of the electrolyte volume is achieved. The entire prediction and compensation process forms a closed-loop control, continuously monitoring the trend of retention changes and dynamically adjusting the compensation intensity to effectively prevent the retention from exceeding the safety boundary, ensuring the continuity and stability of the electrolyzer operation. Through the graded response mechanism, resource waste caused by over-compensation under safe conditions is avoided, while rapid response is ensured under critical conditions to avoid potential risks, achieving refined and intelligent compensation control.

[0051] In one optional implementation, determining the current response level based on the interval in which the prediction margin value falls includes: Set a first margin threshold and a second margin threshold, where the first margin threshold is greater than the second margin threshold, and the first margin threshold and the second margin threshold divide the margin space into three continuous intervals; When the prediction margin value is greater than the first margin threshold, the current response level is determined to be the normal level, and the system is in a safe operating state. When the predicted margin value is less than or equal to the first margin threshold and greater than the second margin threshold, the current response level is determined to be the level of concern, and a gradual pressure control strategy is initiated. When the predicted margin value is less than or equal to the second margin threshold, the current response level is determined to be the emergency level, and the forced decompression protection strategy is activated. A response level migration hysteresis mechanism is established. When the response level migrates from high to low, the first margin threshold and the second margin threshold are used as the judgment boundary. When the response level recovers from low to high, the threshold after adding a preset hysteresis amount is used as the judgment boundary.

[0052] In the intelligent control of gas pressure, to achieve precise hierarchical management of the margin space, it is necessary to scientifically divide the margin space and establish a dynamic response mechanism. Specifically, two key margin thresholds are set to achieve a quantitative assessment of the pressure safety situation. The first margin threshold is used to define the transition zone from a safe operating state to a state requiring attention. The setting of this threshold needs to comprehensively consider the pipeline design pressure, historical operating data, and equipment pressure-bearing capacity. In practical engineering applications, the first margin threshold is usually set between 60% and 70% of the design margin space to ensure sufficient warning time before the pressure approaches the critical state. The second margin threshold serves as the critical point for triggering emergency intervention measures. Its value should be significantly lower than the first margin threshold, generally set within the range of 30% to 40% of the design margin space, to ensure that effective mandatory protection measures can be implemented before the pressure truly threatens system safety.

[0053] These two thresholds form three continuous and non-overlapping intervals in the margin space, corresponding to different system operating states. When the predicted margin value calculated based on the prediction model falls within a different interval, the system automatically triggers the corresponding response mechanism. When the predicted margin value is greater than the first margin threshold, it indicates that there is sufficient buffer space between the current pipeline pressure and the safety boundary, and the response level is determined to be at the normal level. In this state, the pressure control system operates according to the established daily operating mode, only needing to perform routine pressure monitoring and data acquisition tasks. The pressure regulating equipment maintains its normal pressure control strategy without requiring special intervention. Energy consumption is optimally configured in this state, meeting user gas demand while avoiding unnecessary frequent equipment operation.

[0054] When the predicted margin value drops to less than or equal to the first margin threshold but still greater than the second margin threshold, the system enters a transitional zone requiring close monitoring, at which point the response level is determined to be of concern. At this level, a gradual pressure control strategy is immediately initiated. This strategy gradually improves the system's margin status through phased, small-amplitude pressure adjustments, avoiding sudden, large-amplitude pressure adjustments that could impact user gas supply. Specifically, a detailed analysis of the current pressure distribution is conducted first to identify key nodes or pipe sections causing the margin value to decrease. For the identified key locations, pressure is fine-tuned according to a preset adjustment step size, with each adjustment controlled between 2% and 5% of the design pressure. After each pressure adjustment operation, the trend of the margin value is monitored in real time, and the need for further adjustment or the direction of adjustment is determined based on the monitoring results. This gradual strategy effectively prevents pressure from developing into dangerous areas while ensuring gas supply stability.

[0055] If the predicted margin value further decreases to less than or equal to the second margin threshold, it indicates that the system has approached or reached a critical safety state. At this point, the response level is immediately determined to be emergency, and a forced pressure reduction protection strategy is simultaneously activated. This strategy prioritizes reducing pipeline pressure as quickly as possible by rapidly opening pressure-reducing valves, activating auxiliary pressure relief devices, or temporarily shutting down some gas supply branches to quickly bring the pressure back to a safe range. During the emergency response, priority is given to ensuring the gas supply safety of core areas and critical users. For non-critical loads, flow restriction or gas supply suspension measures can be temporarily implemented. All emergency operations must be recorded in detail, including the operation time, operation content, and pressure response data, to provide a basis for subsequent accident analysis and system optimization.

[0056] To prevent system oscillations caused by frequent jumps in response levels near threshold boundaries, establishing a response level migration hysteresis mechanism is crucial. The core of this mechanism lies in using different criteria for escalation and de-escalation. When the pressure situation deteriorates and the response level migrates from high to low, the first and second margin thresholds are strictly used as the judgment boundaries to ensure timely escalation of response measures when a downward pressure trend appears, avoiding delays in intervention. Conversely, when the pressure situation improves and the response level recovers from low to high, the threshold with an added preset hysteresis is used as the judgment boundary. The preset hysteresis needs to be set in conjunction with the pipeline network's inertial characteristics and pressure fluctuation patterns, generally taking 8% to 15% of the original threshold. For example, when the system recovers from an emergency level to a concern level, de-escalation is not immediately triggered when the predicted margin value exceeds the second margin threshold; instead, the level reduction operation is only executed after the predicted margin value reaches the sum of the second margin threshold and the hysteresis. This asymmetric judgment mechanism forms a buffer zone near the threshold, effectively suppressing frequent level switching caused by small pressure fluctuations.

[0057] In actual operation, the hysteresis mechanism also needs to be combined with the time dimension. Even if the prediction margin value reaches the degradation condition, the level transition must be confirmed only after the state has lasted for a certain period of time. This duration can be set to 3 to 10 minutes based on the historical pressure fluctuation cycle. Through the dual constraints of spatial and temporal thresholds, the switching of response levels is ensured to be both sensitive and stable, achieving efficient and intelligent control of gas pipeline network pressure and guaranteeing the long-term safe and stable operation of the system.

[0058] In one optional implementation, the flow ratio between the irrigation input pipeline and the irrigation output pipeline is adjusted collaboratively based on the composite control quantity, including: The composite control quantity is decomposed into a pressure maintenance component and a volume balance component. The pressure maintenance component is used to adjust the pressure level in the surgical cavity, and the volume balance component is used to adjust the total amount of irrigation fluid in the surgical cavity. The main regulating pipeline is determined based on the deviation direction between the current intraoperative pressure value and the center value of the preset steady-state range. When the intraoperative pressure value is higher than the center value, the irrigation output pipeline is determined as the main regulating pipeline, and when the intraoperative pressure value is lower than the center value, the irrigation input pipeline is determined as the main regulating pipeline. Calculate the flow ratio between the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation, and apply pipeline flow constraints to the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation respectively to ensure that the flow regulation of each pipeline does not exceed the maximum flow capacity and minimum flow capacity of the corresponding pipeline. The constrained flow rate adjustment is simultaneously sent to the actuators of both the irrigation input pipeline and the irrigation output pipeline to achieve coordinated regulation of the two pipelines.

[0059] After obtaining the composite control quantity, it needs to be converted into specific flow regulation commands for the irrigation input and output lines. The composite control quantity essentially reflects two control objectives: first, maintaining the intraoperative cavity pressure near the desired level; and second, balancing the total volume of irrigation fluid within the cavity. To achieve these two objectives, the composite control quantity is decomposed into components.

[0060] The pressure maintenance component primarily responds to the degree to which the intraoperative cavity pressure deviates from the preset steady-state range center value. When the pressure deviation is large, this component has a higher weight, driving rapid adjustment of the tubing flow to correct the pressure deviation. The volume balance component focuses on the net inflow or outflow trend of the irrigation fluid within the intraoperative cavity. When a continuous increase or decrease in the intraoperative cavity volume is detected, this component comes into play, stabilizing the cavity volume by adjusting the difference between the input and output flow rates. The weighting of these two components is dynamically adjusted according to the magnitude of the current pressure deviation and the rate of volume change. When the pressure is close to the center value, the weight of the volume balance component increases to ensure long-term stability; when the pressure deviation is significant, the weight of the pressure maintenance component increases to achieve rapid response.

[0061] After decomposing the pressure into two components, a control strategy needs to be determined. Based on the relationship between the current intraoperative pressure and the preset steady-state range center value, it is determined which pipeline should be adjusted first. When the intraoperative pressure is higher than the center value, it indicates that the intraoperative pressure is too high. In this case, the flow rate of the irrigation output pipeline should be increased or the flow rate of the irrigation input pipeline should be decreased. To achieve a faster adjustment effect, the irrigation output pipeline is designated as the primary control pipeline, undertaking the main flow rate adjustment task. Correspondingly, the irrigation input pipeline acts as the secondary control pipeline, performing auxiliary flow rate adjustments. When the intraoperative pressure is lower than the center value, it indicates that the intraoperative pressure is too low. In this case, the flow rate of the irrigation input pipeline should be increased or the flow rate of the irrigation output pipeline should be decreased. The irrigation input pipeline is designated as the primary control pipeline, and the irrigation output pipeline acts as the secondary control pipeline. This dynamic switching mechanism between primary and secondary pipelines ensures that the adjustment process always proceeds in the optimal direction for pressure correction.

[0062] After determining the main regulating pipeline, the flow regulation ratio between the two pipelines is calculated. The flow regulation of the main regulating pipeline is directly related to the pressure maintenance component, and the flow adjustment range is determined based on the pressure deviation. The flow regulation of the secondary regulating pipeline comprehensively considers the volume balance component and the need for coordinated regulation. It must cooperate with the main regulating pipeline to complete pressure adjustment while avoiding excessive fluctuations in the surgical cavity volume. In the specific ratio calculation, the flow regulation of the main regulating pipeline is set as the pressure maintenance component multiplied by the main regulation coefficient, and the flow regulation of the secondary regulating pipeline is set as the volume balance component multiplied by the secondary regulation coefficient. The main regulation coefficient is usually set to 0.6 to 0.9, and the secondary regulation coefficient is set to 0.1 to 0.4, with their sum close to 1.0, ensuring that the composite control is fully utilized. In practical applications, the coefficients are fine-tuned according to the surgical type and surgical cavity characteristics. For volume-sensitive surgical scenarios, the secondary regulation coefficient is appropriately increased to enhance volume stability.

[0063] After calculating the initial flow regulation, pipeline flow constraints must be applied. Each pipeline has its maximum and minimum flow capacity limits. The maximum flow capacity is limited by the pump's maximum output power, the pipeline's inner diameter, and the viscosity of the irrigation fluid. The minimum flow capacity considers the pump's starting flow threshold and the continuity requirements of the fluid within the pipeline. For the irrigation input pipeline, the maximum flow capacity is typically in the range of 100 to 300 ml / min, and the minimum flow capacity is no less than 5 ml / min. For the irrigation output pipeline, the maximum flow capacity is affected by the negative pressure suction intensity, generally in the range of 80 to 250 ml / min, and the minimum flow capacity is also no less than 5 ml / min. When applying constraints, first check whether the flow regulation of the main regulating pipeline exceeds the pipeline's capacity range. If it exceeds the maximum flow capacity, limit it to the maximum flow capacity value; if it is lower than the minimum flow capacity, limit it to the minimum flow capacity value or zero. The same treatment is applied to the flow regulation of the secondary regulating pipelines to ensure that the flow regulation of both pipelines is within the physically feasible range.

[0064] During the constraint process, the rate of change of the flow regulation also needs to be considered. Excessive flow rate changes can cause drastic fluctuations in intraoperative pressure, affecting surgical stability. Therefore, an upper limit is set for the rate of change of flow, typically limited to 50 ml / min / s. When the calculated rate of change of the flow regulation relative to the current flow exceeds this upper limit, the flow regulation is smoothed using first-order filtering or ramp limiting methods to gradually adjust the flow to the target value at a controlled rate. This rate of change constraint is particularly important in scenarios with rapid pressure fluctuations, as it can prevent overshoot or oscillations in the control system.

[0065] After constraint processing, the constrained flow regulation is converted into control commands for the actuators. The actuators in the irrigation input pipeline are typically peristaltic pumps or plunger pumps, and the control commands are motor speed or drive voltage. The required motor speed is calculated based on the correspondence between the flow regulation and the pump characteristic curve. The actuators in the irrigation output pipeline are negative pressure suction pumps or solenoid valve-controlled drainage devices, and the control commands are negative pressure setpoints or valve opening degrees. The negative pressure magnitude or valve opening percentage is determined based on the mapping relationship between the flow regulation and suction capacity. To ensure synchronized flow adjustments in both pipelines, a unified command issuance time is used, sending both control commands simultaneously to their respective actuators within the same control cycle.

[0066] The synchronous command issuance mechanism also includes command priority settings and failure protection. When the main regulating pipeline actuator malfunctions or responds abnormally, it immediately switches to single-pipeline regulation mode, with the slave regulating pipeline undertaking all regulation tasks. Simultaneously, an alarm is triggered to prompt surgical personnel to check the equipment status. When the slave regulating pipeline actuator malfunctions, the main regulating pipeline continues to operate, but the regulation amplitude is appropriately reduced to avoid imbalance in the surgical cavity volume caused by single-pipeline regulation. Under normal operating conditions, at the end of each control cycle, the actual flow feedback values ​​of the two pipelines are collected and compared with the flow regulation amount to calculate the execution deviation. If the execution deviation exceeds a threshold, the flow regulation amount is compensated and corrected in the next control cycle to ensure the accuracy and robustness of closed-loop control.

[0067] A second aspect of the present invention provides a multi-sensor pressure balancing control system for bladder irrigation, comprising: The pressure gradient unit is used to synchronously acquire the input pressure value and the output pressure value through the pressure sensing unit installed in the irrigation input pipeline and the irrigation output pipeline, and to calculate the pressure gradient value between the two. The fuzzy control unit is used to construct an adaptive fuzzy control mechanism based on the pressure gradient value. The gradient deviation between the pressure gradient value and the target gradient value, as well as the time derivative of the gradient deviation value, are used as dual-input fuzzy variables. The dual-input fuzzy variables are mapped to the fuzzy domain through a membership function. Fuzzy inference is performed according to a preset fuzzy rule base, and the flow rate adjustment increment is output after defuzzification processing. The pulse width modulation unit is used to generate a pulse width modulation signal according to the flow rate adjustment increment, drive the proportional regulating valve set in the irrigation input pipeline to adjust its valve opening, and synchronously regulate the input flow rate of the irrigation fluid. The volume estimation unit is used to establish a dynamic estimation mechanism for the surgical cavity volume. It calculates the amount of irrigation fluid retained in the surgical cavity based on the difference between the cumulative input flow rate of the irrigation input pipeline and the cumulative output flow rate of the irrigation output pipeline. The retained amount is compared with a preset volume safety boundary. When the retained amount approaches the volume safety boundary, a volume compensation coefficient is generated. The composite control unit is used to weight and fuse the volume compensation coefficient and the flow rate adjustment increment to obtain a composite control quantity, and based on the composite control quantity, coordinately adjust the flow rate ratio of the irrigation input pipeline and the irrigation output pipeline to maintain the intraoperative cavity pressure within a preset steady-state range.

[0068] A third aspect of the present invention provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0069] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0070] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-sensor pressure balance control system for bladder irrigation, characterized in that, include: The pressure gradient unit is used to synchronously acquire the input pressure value and the output pressure value through the pressure sensing unit installed in the irrigation input pipeline and the irrigation output pipeline, and to calculate the pressure gradient value between the two. The fuzzy control unit is used to construct an adaptive fuzzy control mechanism based on the pressure gradient value. The gradient deviation between the pressure gradient value and the target gradient value, as well as the time derivative of the gradient deviation value, are used as dual-input fuzzy variables. The dual-input fuzzy variables are mapped to the fuzzy domain through a membership function. Fuzzy inference is performed according to a preset fuzzy rule base, and the flow rate adjustment increment is output after defuzzification processing. The pulse width modulation unit is used to generate a pulse width modulation signal according to the flow rate adjustment increment, drive the proportional regulating valve set in the irrigation input pipeline to adjust its valve opening, and synchronously regulate the input flow rate of the irrigation fluid. The volume estimation unit is used to establish a dynamic estimation mechanism for the surgical cavity volume. It calculates the amount of irrigation fluid retained in the surgical cavity based on the difference between the cumulative input flow rate of the irrigation input pipeline and the cumulative output flow rate of the irrigation output pipeline. The retained amount is compared with a preset volume safety boundary. When the retained amount approaches the volume safety boundary, a volume compensation coefficient is generated. The composite control unit is used to weight and fuse the volume compensation coefficient and the flow rate adjustment increment to obtain a composite control quantity, and based on the composite control quantity, coordinately adjust the flow rate ratio of the irrigation input pipeline and the irrigation output pipeline to maintain the intraoperative cavity pressure within a preset steady-state range.

2. The system according to claim 1, characterized in that, The fuzzy control unit is also used for: The pressure gradient value is sampled by a sliding window to obtain a pressure gradient sequence. The pressure gradient mean and pressure gradient variance are calculated based on the pressure gradient sequence. The pressure gradient variance is used to determine whether the current pressure fluctuation state belongs to a steady state mode or a disturbance mode. In the steady-state mode, the dual-input fuzzy variables are quantized using a first universe of discourse range, and in the perturbation mode, the dual-input fuzzy variables are quantized using a second universe of discourse range, wherein the second universe of discourse range is larger than the first universe of discourse range. The input quantization factor is dynamically calculated based on the ratio of the pressure gradient variance to the preset variance benchmark value. The gradient deviation value and the corresponding gradient deviation change rate are then multiplied by the input quantization factor and mapped to the corresponding domain range. A Gaussian membership function set is constructed for the quantized dual-input fuzzy variables. Each language variable level corresponds to a Gaussian membership function, and the Gaussian membership functions of adjacent language variable levels maintain a preset overlap. Based on the fuzzy rule base, inference operations are performed to obtain the fuzzy output. The corresponding output scaling factor is selected according to the pressure fluctuation state to scale the defuzzification result and generate the flow regulation increment.

3. The system according to claim 2, characterized in that, The fuzzy control unit is also used for: The number of language variable levels is determined based on the domain range of the dual-input fuzzy variables. The domain range is then uniformly divided into multiple sub-intervals corresponding to the number of language variable levels. The center point of each sub-interval is used as the peak center of the Gaussian membership function of the corresponding language variable level. The width parameter of the Gaussian membership function is calculated based on the preset overlap between adjacent linguistic variable levels, and the width parameter determines the decay rate of the Gaussian membership function. A first Gaussian membership function group is constructed for the gradient deviation value. Each Gaussian membership function in the first Gaussian membership function group shares the same width parameter and the peak centers are arranged sequentially according to the center points of the sub-intervals. A second Gaussian membership function set is constructed for the gradient deviation change rate, and the construction method of the second Gaussian membership function set is the same as that of the first Gaussian membership function set.

4. The system according to claim 1, characterized in that, The pulse width modulation unit is also used for: The flow rate adjustment increment is rate-limited. The flow rate difference between the current flow rate adjustment increment and the previous cycle flow rate adjustment increment is calculated. When the flow rate difference exceeds the preset increment change limit, the flow rate adjustment increment is corrected to the limit boundary to generate a rate-limited flow rate adjustment increment. The rate-limited flow regulation increment is reverse-compensated based on the flow characteristic curve of the proportional control valve to eliminate the influence of the valve's inherent nonlinear characteristics on the flow control accuracy. The target duty cycle is calculated based on the compensated flow regulation increment, and the pulse width modulation signal is generated by combining it with the preset modulation frequency. The duty cycle of the pulse width modulation signal is consistent with the target duty cycle. A jitter component is superimposed on the pulse width modulation signal, the amplitude of which is less than the duty cycle resolution; the pulse width modulation signal after superimposing the jitter component is amplified and then output to the proportional control valve to drive the proportional control valve to adjust its valve opening.

5. The system according to claim 1, characterized in that, The volume estimation unit is also used for: Linear fitting is performed on the continuously collected retention data to obtain the slope of retention change. Based on the current retention and the slope of retention change, the predicted retention after a preset time window is predicted. The difference between the preset volume safety boundary and the predicted retention amount is used as the prediction margin value; The current response level is determined based on the range in which the predicted margin value falls, and the response level includes normal level, attention level and emergency level. Different compensation coefficient generation strategies are configured for different response levels. The regular level corresponds to a unit compensation coefficient, the attention level corresponds to a gradual compensation coefficient based on linear interpolation of the prediction margin value, and the emergency level corresponds to an enhanced compensation coefficient based on exponentially increasing prediction margin value. The compensation coefficient generation strategy corresponding to the current response level is applied to the prediction margin value, and the volume compensation coefficient is output.

6. The system according to claim 5, characterized in that, The volume estimation unit is also used for: Set a first margin threshold and a second margin threshold, where the first margin threshold is greater than the second margin threshold, and the first margin threshold and the second margin threshold divide the margin space into three continuous intervals; When the prediction margin value is greater than the first margin threshold, the current response level is determined to be the normal level, and the system is in a safe operating state. When the predicted margin value is less than or equal to the first margin threshold and greater than the second margin threshold, the current response level is determined to be the level of concern, and a gradual pressure control strategy is initiated. When the predicted margin value is less than or equal to the second margin threshold, the current response level is determined to be the emergency level, and the forced decompression protection strategy is activated. A response level migration hysteresis mechanism is established. When the response level migrates from high to low, the first margin threshold and the second margin threshold are used as the judgment boundary. When the response level recovers from low to high, the threshold after adding a preset hysteresis amount is used as the judgment boundary.

7. The system according to claim 1, characterized in that, The composite control unit is also used for: The composite control quantity is decomposed into a pressure maintenance component and a volume balance component. The pressure maintenance component is used to adjust the pressure level in the surgical cavity, and the volume balance component is used to adjust the total amount of irrigation fluid in the surgical cavity. The main regulating pipeline is determined based on the deviation direction between the current intraoperative pressure value and the center value of the preset steady-state range. When the intraoperative pressure value is higher than the center value, the irrigation output pipeline is determined as the main regulating pipeline, and when the intraoperative pressure value is lower than the center value, the irrigation input pipeline is determined as the main regulating pipeline. Calculate the flow ratio between the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation, and apply pipeline flow constraints to the main regulating pipeline flow regulation and the slave regulating pipeline flow regulation respectively to ensure that the flow regulation of each pipeline does not exceed the maximum flow capacity and minimum flow capacity of the corresponding pipeline. The constrained flow rate adjustment is simultaneously sent to the actuators of both the irrigation input pipeline and the irrigation output pipeline to achieve coordinated regulation of the two pipelines.

Images