A remote fluid switching type rainwater multi-parameter dynamic online analysis system

Abstract

The present application relates to the field of physical test technology, and provides a remote fluid switching type rainwater multi-parameter dynamic online analysis system, which comprises a remote multi-channel switching valve, a multi-parameter sensor array and a closed-loop control assembly; the closed-loop control assembly collects fluid flow rate mutation signals at switching moments of the remote multi-channel switching valve and transient response signals of the multi-parameter sensor array in real time, calculates fluid dynamics disturbance factors at the switching moments; and performs convolution operation on the fluid dynamics disturbance factors and baseline drift amounts of sensing signals of the multi-parameter sensor array to generate dynamic compensation coefficients. The system of the present application overcomes the measurement baseline drift defect in the remote fluid switching process, eliminates the interference of flow rate mutation at the switching moments on the sensor response, reduces flush water consumption in the fluid switching process, eliminates the disturbance of micro-bubble aggregation in the pipeline on the fluid flow state, and guarantees the measurement stability of the multi-parameter online analysis system in complex physical environments.

Description

Technical Field

[0001] This invention relates to the field of physical testing technology, specifically to a remote fluid switching type rainwater multi-parameter dynamic online analysis system. Background Technology

[0002] In the field of dynamic online analysis of rainwater multi-parameters, conventional technologies employ a system architecture combining a remote multi-channel switching valve with a multi-parameter sensor array to centrally monitor rainwater physicochemical indicators. The remote multi-channel switching valve is responsible for sequentially switching rainwater from different sources to the measurement channel. Existing control methods rely on preset fixed timing logic. The control system sends equal-width drive pulses to the switching valve's stepper motor at set time intervals, driving the valve core to rotate at a constant angular velocity to the target channel. After the valve core reaches the target position, the system maintains fluid flow for a preset fixed flushing time. Once the flushing time is exhausted, the sensor array is triggered to perform data acquisition. This approach treats fluid switching as a simple mechanical displacement process, focusing only on the physical position of the valve core and the fixed flushing duration, without considering the dynamic physical characteristics of fluid movement within the pipeline.

[0003] Under fixed-sequence control operation, the sensor signal processing stage is unidirectionally decoupled from the fluid switching action. The action of the remote multi-channel switching valve causes a sudden change in the cross-sectional area and flow direction of the fluid channel, resulting in abrupt changes in the flow velocity within the pipeline and generating fluid dynamic disturbances. This disturbance affects the sensor sensing end face, causing a baseline shift in the output signal. Existing technologies address this baseline shift using either a fixed-delay waiting strategy in the time domain or conventional low-pass filtering at the signal processing end. Fixed-delay waiting extends the waiting time before data acquisition, hoping for natural fluid stabilization; low-pass filtering attempts to filter out high-frequency disturbance signals. The core logic of both methods passively waits for or processes the sensor signal after the fluid switching action is completed. There is no interactive feedback between the fluid switching mechanical action parameters and the sensor signal response state, and the flushing cycle is set to a static constant.

[0004] The fluid dynamics within stormwater drainage networks are complex and highly variable. Fluid viscosity, velocity, and boundary layer conditions differ under varying rainfall intensities and network confluence. Fixed-sequence control of switching actions and fixed-delay signal processing cannot adapt to these dynamic changes. When fluid velocity changes drastically, the fixed flushing wait time is insufficient to stabilize the sensor signal, resulting in persistent residual fluid dynamic disturbances and dynamic drift of the acquired data baseline. Conversely, when velocity changes are gradual, the fixed wait time leads to water and time consumption. Current technologies lack a means to establish a closed-loop feedback relationship between the physical disturbances caused by fluid switching and the sensor signal response. This mismatch between remote fluid switching actions and the sensor signal stabilization process leads to the core technical problem of dynamic drift of multi-parameter measurement baselines caused by remote fluid switching. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the above-mentioned defects of the prior art and provide a remote fluid switching type rainwater multi-parameter dynamic online analysis system that overcomes the measurement baseline drift defect during the remote fluid switching process, eliminates the interference of sudden flow velocity change at the moment of switching on the sensor response, reduces flushing water consumption during the fluid switching process, eliminates the disturbance of fluid flow state caused by microbubble accumulation in the pipeline, and ensures the measurement stability of the multi-parameter online analysis system in complex physical environments.

[0006] The technical solution adopted by the present invention to solve its technical problem is as follows: a remote fluid switching rainwater multi-parameter dynamic online analysis system, including a remote multi-channel switching valve, a multi-parameter sensor array and a closed-loop control component; The closed-loop control component collects the fluid velocity change signal at the moment of switching of the remote multi-channel switching valve and the transient response signal of the multi-parameter sensor array in real time, and calculates the fluid dynamic disturbance factor at the moment of switching. The fluid dynamics disturbance factor is convolved with the baseline drift of the sensing signal of the multi-parameter sensor array to generate dynamic compensation coefficients. The closed-loop control component adjusts the step drive pulse width and switching interval timing of the remote multi-channel switching valve in reverse based on the dynamic compensation coefficient, so that the fluid switching action and the sensor signal stabilization process are deeply coupled.

[0007] Preferably, the closed-loop control component extracts the Doppler frequency shift characteristics of the pressure fluctuation in the internal flow channel of the remote multi-channel switching valve body as the fluid velocity change signal, and simultaneously extracts the phase delay difference of the output electrical signal of the multi-parameter sensor array as the transient response signal; The closed-loop control component constructs a timing alignment window based on the pipeline space span between the remote multi-channel switching valve and the multi-parameter sensor array. It performs cross-correlation calculation on the Doppler frequency shift feature and the phase delay difference within the timing alignment window, and extracts the transient coupling feature vector at the moment of fluid switching as the input parameter of the fluid dynamics disturbance factor.

[0008] Preferably, the closed-loop control component dynamically adjusts the width of the timing alignment window based on the peak offset of the cross-correlation calculation, and establishes a sliding step size within the timing alignment window; The Doppler frequency shift feature and the phase delay difference are compared frame by frame within the time alignment window according to the sliding step size. The maximum mutual information between the Doppler frequency shift feature and the phase delay difference is extracted as the principal component of the transient coupling feature vector. Redundant components in the transient coupling feature vector that are below the information entropy threshold are removed.

[0009] Preferably, the closed-loop control component constructs a nonlinear fluid dynamics state equation based on the viscosity coefficient of the current rainwater fluid and the elastic modulus of the channel wall; The fluid velocity change signal and the transient response signal are input into the nonlinear fluid dynamics state equation to solve for the fluid boundary layer separation index and eddy current dissipation rate. The closed-loop control component performs nonlinear fitting between the fluid boundary layer separation index and the eddy current dissipation rate to generate the fluid dynamic disturbance factor characterizing the mechanical impulse disturbance at the moment of fluid switching.

[0010] Preferably, the closed-loop control component calculates the Reynolds number in the pipeline at the front end of the remote multi-channel switching valve in real time. When the Reynolds number crosses the critical threshold between laminar and turbulent flow, it triggers the boundary condition reconstruction of the nonlinear fluid dynamics state equation. The closed-loop control component introduces an eddy viscosity correction term under turbulent boundary conditions and a friction loss correction term under laminar boundary conditions. Based on the eddy viscosity correction term or the friction loss correction term, it adaptively switches the solution step size of the nonlinear fluid dynamics state equation and outputs the corrected fluid dynamics disturbance factor.

[0011] Preferably, the closed-loop control component converts the hydrodynamic disturbance factor and the baseline drift of the sensing signal to the frequency domain space, and calculates the coherence index between the frequency domain distribution characteristics of the hydrodynamic disturbance factor and the frequency domain distribution characteristics of the baseline drift of the sensing signal in the frequency domain space; Based on the coherence index, steady-state drift components and transient disturbance components are separated. Differentiated weighting coefficients are assigned to the steady-state drift components and the transient disturbance components, and inverse frequency domain transformation is performed to generate the dynamic compensation coefficients containing multi-band compensation vectors.

[0012] Preferably, the closed-loop control component performs a difference calculation between the coherence index and a preset coherence benchmark to obtain the coherence attenuation gradient; The allocation ratio of the differentiated weight coefficients is dynamically adjusted according to the rate of change of the coherence decay gradient. When the rate of change of the coherence decay gradient exceeds the steady-state range, the weight coefficient ratio of the transient disturbance component is increased. The multi-band compensation vector is regenerated based on the adjusted weight coefficient ratio, and the multi-band compensation vector is fed back to the input of the inverse frequency domain conversion for iterative update.

[0013] Preferably, the closed-loop control component parses the multi-band compensation vector in the dynamic compensation coefficient and maps the multi-band compensation vector to a non-uniform subdivision drive pulse sequence of the stepper motor in the remote multi-channel switching valve; The non-uniform subdivision driving pulse sequence includes pulse step angles arranged in a non-arithmetic sequence; The closed-loop control component adjusts the duty cycle combination of the non-uniform subdivision drive pulse sequence according to the amplitude of the dynamic compensation coefficient, so as to control the valve core of the remote multi-channel switching valve to perform variable acceleration rotation and dynamically modify the dwell time node of the switching interval sequence.

[0014] Preferably, it also includes a photoelectric coupling monitoring array and a piezoelectric micro-frequency vibration component arranged along the outlet pipeline of the remote multi-channel switching valve; The photoelectric coupling monitoring array acquires an image of the light transmittance distribution of the fluid flowing through the pipeline, and identifies discrete low transmittance regions in the light transmittance distribution image as microbubble distribution features. The closed-loop control component calculates the microbubble aggregation density and spatial distribution centroid based on the microbubble distribution characteristics, generates a resonant frequency command based on the microbubble aggregation density and the spatial distribution centroid, and drives the piezoelectric micro-frequency vibration component to generate a high-frequency micro-amplitude oscillation deformation at the pipeline position corresponding to the spatial distribution centroid that matches the resonant frequency command.

[0015] Preferably, it further includes a temperature distribution gradient collector arranged between the remote multi-channel switching valve and the multi-parameter sensor array; The temperature distribution gradient collector acquires the temperature sequence at multiple points along the axial direction of the pipeline and constructs a thermal gradient field vector along the fluid flow direction. The closed-loop control component performs orthogonal decoupling operation on the thermal gradient field vector and the hydrodynamic disturbance factor, separates the thermally induced baseline drift component affected by the thermal gradient field vector, subtracts the thermally induced baseline drift component from the baseline drift of the sensing signal, and performs the convolution operation on the remaining drift component and the hydrodynamic disturbance factor.

[0016] The beneficial effects of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention are as follows: (1) In the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention, the closed-loop control component collects the fluid velocity change signal and the transient response signal of the multi-parameter sensor array at the moment of switching of the remote multi-channel switching valve in real time, calculates the fluid dynamic disturbance factor, and performs convolution operation on the fluid dynamic disturbance factor and the baseline drift of the sensor signal to generate a dynamic compensation coefficient. Based on the dynamic compensation coefficient, the step drive pulse width and switching interval timing of the remote multi-channel switching valve are adjusted in reverse to make the fluid switching action and the sensor signal stable establishment process deeply coupled. The Doppler frequency shift characteristics of the pressure fluctuation in the flow channel inside the switching valve body and the phase delay difference of the sensor output electrical signal are extracted. Cross-correlation operation is performed within the timing alignment window to extract the transient coupling feature vector. Based on the fluid viscosity coefficient and the elastic modulus of the flow channel wall, the dynamic compensation coefficient is calculated. A nonlinear fluid dynamics state equation is constructed to solve for the fluid boundary layer separation index and eddy current dissipation rate, generating a fluid dynamics disturbance factor. In the frequency domain, the coherence index of the frequency domain distribution characteristics of the fluid dynamics disturbance factor and the frequency domain distribution characteristics of the sensor signal baseline drift is calculated to separate the steady-state drift component and the transient disturbance component. Differentiated weight coefficients are assigned to generate dynamic compensation coefficients containing multi-band compensation vectors. These multi-band compensation vectors are mapped to a non-uniform subdivision drive pulse sequence to control the valve core to perform variable acceleration rotation and dynamically modify the dwell time node of the switching interval. The weight coefficient ratio is dynamically adjusted according to the coherence attenuation gradient change rate to perform iterative updates. These methods overcome the defects of fixed-sequence unidirectional control, eliminate the interference of sudden flow velocity changes during switching on the sensor response, and reduce flushing water consumption during fluid switching. (2) In the remote fluid switching rainwater multi-parameter dynamic online analysis system of the present invention, the closed-loop control component triggers the boundary condition reconstruction of the nonlinear fluid dynamics state equation based on the Reynolds number crossing the critical threshold between laminar and turbulent flow, introduces eddy viscosity correction term or friction loss correction term and adaptively switches the solution step size to adapt to the calculation requirements of disturbance factors under different flow states and improve the computational robustness of the system under complex flow states; by acquiring the light transmittance distribution image of the fluid flowing through the pipeline through the photoelectric coupling monitoring array, the discrete low transmittance region is identified as the microbubble distribution feature, the microbubble aggregation density and spatial distribution centroid are calculated to generate the resonant frequency command, and the piezoelectric micro-frequency vibration component generates a high-frequency micro-amplitude oscillation deformation at the pipeline position corresponding to the spatial distribution centroid that matches the resonant frequency command, thereby eliminating the disturbance of the fluid flow state caused by the aggregation of microbubbles in the pipeline; (3) The remote fluid switching rainwater multi-parameter dynamic online analysis system of the present invention obtains the temperature sequence of multiple points in the axial direction of the pipeline through the temperature distribution gradient collector to construct the thermal gradient field vector along the fluid flow direction. The thermal gradient field vector and the fluid dynamic disturbance factor are orthogonally decoupled to separate the thermally induced baseline drift component affected by the thermal gradient field vector. After deducting the thermally induced baseline drift component from the baseline drift of the sensing signal, the convolution operation is performed to eliminate the intervention of thermal interference factors in the dynamic compensation operation process and ensure the measurement stability of the multi-parameter online analysis system in complex physical environments. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the overall closed-loop dynamic control process of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention. Figure 2 This is a flowchart of the signal feature extraction and time-series alignment coupling operation of an embodiment of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention; Figure 3 This is a flowchart of the signal feature extraction and time-series alignment coupling operation of an embodiment of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention; Figure 4 This is a flowchart of frequency domain coherence analysis and multi-band dynamic compensation coefficient iteration of an embodiment of the remote fluid switching rainwater multi-parameter dynamic online analysis system of the present invention; Figure 5 This is a flowchart of the non-uniform pulse variable acceleration control process of the switching valve in an embodiment of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention. Figure 6 This is a flowchart illustrating the multi-dimensional fluid and environmental interference suppression and compensation process of an embodiment of the remote fluid switching type rainwater multi-parameter dynamic online analysis system of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0019] Example 1: This invention provides a remote fluid switching type rainwater multi-parameter dynamic online analysis system. The system comprises a remote multi-channel switching valve, a multi-parameter sensor array, and a closed-loop control component. The overall equipment is adaptable to scenarios involving continuous switching acquisition and real-time physicochemical parameter analysis of rainwater fluids at multiple points in a rainwater pipe network, and can meet the monitoring needs of dynamic fluid changes under different rainfall conditions. The remote multi-channel switching valve, as the fluid path switching execution structure, has multiple fluid input channels and a single output measurement channel, enabling time-sharing switching of rainwater fluids at different pipe network points. The valve body integrates a stepper drive structure, which receives pulse drive signals to complete valve core rotation, channel opening and closing, and fluid flow on / off actions.

[0020] A multi-parameter sensor array is deployed inside the downstream fluid pipeline of the remote multi-channel switching valve. The array includes conventional rainwater parameter detection units such as pH, conductivity, turbidity, and ammonia nitrogen concentration sensors. The sensing end face of each unit is completely immersed in the flowing rainwater, enabling real-time acquisition of the corresponding physicochemical parameters and electrical signals, forming a continuous sequence of sensor data. The closed-loop control component is the core computing and control unit of the system. It incorporates a high-speed data acquisition module, a floating-point processor, a timing logic controller, and a signal conversion module. This allows for synchronous acquisition, quantization, parameter solving, and closed-loop control of multiple signals, achieving dynamic adaptive control without human intervention throughout the entire process.

[0021] In this embodiment, the core operating logic of the closed-loop control component is to capture the dynamic physical signals and sensor response signals at the moment of fluid switching in real time. It then uses quantization to solve for disturbance characteristics and generate compensation and control parameters, which in turn correct the switching valve's operating sequence and drive parameters, achieving dynamic matching between the fluid switching action and the sensor signal stabilization process. Specifically, the closed-loop control component uses a built-in high-speed signal acquisition unit to synchronously acquire the fluid velocity mutation signal at the moment of switching of the remote multi-channel switching valve, as well as the transient response signal synchronously output by the multi-parameter sensor array, at a microsecond-level sampling frequency. The acquisition timing of these two types of signals is strictly aligned with the start and stop of the switching valve core action, avoiding invalid data acquisition during the steady-state fluid phase. The fluid velocity mutation signal characterizes the instantaneous change in flow velocity caused by changes in the cross-sectional area and flow direction of the pipeline fluid during valve core switching, while the transient response signal characterizes the instantaneous offset, fluctuation, and delay characteristics of the output electrical signal of the sensor array after being disturbed by the fluid.

[0022] like Figure 1As shown, after acquiring two types of raw signals, the closed-loop control component calculates the fluid dynamic disturbance factor at the moment of switching using a built-in fluid dynamics quantification algorithm. This factor is a comprehensive characteristic parameter that quantifies the degree of disturbance to the pipeline fluid flow field, pressure field, and sensor detection environment caused by the fluid switching mechanical action. It can accurately characterize the intensity, duration, and range of the instantaneous disturbance. After solving for the disturbance factor, the closed-loop control component retrieves the baseline drift of the real-time sensing signal from the multi-parameter sensor array. This drift is the cumulative deviation between the steady-state detection reference value and the real-time output value of the sensor, including all drift components such as dynamic drift caused by fluid disturbance, inherent static drift of the equipment, and indirect drift caused by environmental interference.

[0023] Furthermore, the closed-loop control component performs a standard two-dimensional convolution operation on the fluid dynamics disturbance factor and the baseline drift of the sensor signal. Through pixel-level data association and weight superposition, it generates dynamic compensation coefficients that adapt to the current fluid disturbance state. These coefficients are multi-dimensional compensation parameters that include time-domain and amplitude features, and can specifically offset the baseline offset problem of sensor signals caused by fluid switching disturbances. The core logic of the convolution operation is to establish a spatial mapping relationship between the time-series matrix of the disturbance factor and the time-series matrix of the baseline drift, achieving precise coupling and matching of disturbance features and drift features, thus overcoming the single-parameter deficiency of traditional fixed-parameter compensation.

[0024] The convolution operation used in this embodiment is quantized using the following formula: ; Where C(n) represents the dynamic compensation coefficient corresponding to the nth time node; F(k) represents the fluid dynamic disturbance factor at the kth sampling time; D(nk) represents the baseline drift of the sensor signal corresponding to the time offset nk; and N is the total number of signal sampling points corresponding to a single operation, which is adaptively set according to the pipeline fluid response time, ranging from 200 to 1000 points. This formula correlates and matches the fluid disturbance characteristics at different times with the baseline drift characteristics through point-by-point time-series superposition, accurately extracting the core drift component caused by the disturbance and avoiding interference from irrelevant noise signals in the generation of the compensation coefficient.

[0025] After generating the dynamic compensation coefficients, the closed-loop control component maps these multi-dimensional parameters to the drive and control parameters of the remote multi-channel switching valve, and specifically adjusts the stepper drive pulse width and switching interval timing. Adjusting the pulse width changes the energy output of the stepper motor's single-step drive, thereby controlling the instantaneous angular velocity of the valve core rotation. Adjusting the switching interval timing dynamically corrects the fluid residence time and steady-state establishment waiting time after the fluid channel switching, ensuring that the mechanical motion rhythm of fluid switching is perfectly matched with the physical process of sensor signal attenuation and stabilization, achieving deep coupling between the fluid switching action and the sensor signal stabilization process. This control logic breaks away from the traditional unidirectional control mode with fixed timing and fixed pulse parameters, forming a closed-loop feedback through real-time signal acquisition and quantization calculations, adapting to the dynamic disturbance differences under different rainwater fluid conditions.

[0026] This embodiment, as a core basic embodiment, fully covers all the technical features of the system's core architecture. Through a complete closed-loop logic of real-time signal acquisition, disturbance factor quantization, convolution compensation operation, and dynamic correction of driving parameters, it solves the technical problems of fluid disturbance and sensor signal mismatch and measurement baseline drift caused by traditional fixed timing control.

[0027] In a preferred embodiment, based on the system architecture of the aforementioned embodiments, the signal extraction and timing alignment operation logic of the closed-loop control component is further refined to adapt to the requirement of accurate extraction of disturbance features under complex fluid switching conditions. In this embodiment, the closed-loop control component abandons the traditional single flow velocity signal acquisition method and adopts a multi-dimensional feature extraction mechanism to obtain the core feature components of the fluid flow velocity change signal and transient response signal. Specifically, the closed-loop control component uses a high-frequency pressure sensing unit deployed on the inner wall of the flow channel of the remote multi-channel switching valve to perform spectral analysis on the continuous pressure fluctuation time-domain signal, extract the Doppler frequency shift feature during the pressure fluctuation process, and use this frequency shift feature as the core characterization parameter of the fluid flow velocity change signal.

[0028] When a sudden change in fluid velocity occurs inside the valve body, the fluid medium undergoes relative motion with respect to the valve body pressure detection unit. This pressure fluctuation signal generates a corresponding Doppler frequency shift. The magnitude of the frequency shift is positively correlated with the instantaneous change in fluid velocity, and the fluctuation frequency is directly related to the duration and rate of the velocity change. Therefore, the dynamic change in fluid velocity at the moment of switching can be accurately quantified using Doppler frequency shift characteristics. Compared to traditional single-point velocity detection methods, this feature extraction method avoids detection errors caused by local turbulence and microbubble interference in the pipeline. Simultaneously, the closed-loop control component synchronously acquires the analog electrical signals output by the multi-parameter sensor array. It compares the phase of the steady-state reference electrical signal with the instantaneous electrical signal at the moment of switching, calculates the phase delay difference between the two sets of signals, and uses this phase delay difference as the transient response signal of the sensor array. When the sensor is affected by fluid dynamic disturbances, the fluid adhesion state and pressure state at the sensing end face change instantaneously, causing a phase shift in the output electrical signal. The magnitude of the phase delay difference accurately characterizes the sensor's response degree and response rate to fluid disturbances.

[0029] Furthermore, the closed-loop control component constructs an adaptive timing alignment window based on the pipeline space span parameters between the remote multi-channel switching valve and the multi-parameter sensor array. The pipeline space span parameters include the straight-line distance from the switching valve outlet to the sensor detection area, the number of pipe bends, and the pipe diameter. Different pipeline space parameters will cause differences in the transmission time of fluid disturbances from the switching valve to the sensor. The timing alignment window is a time-domain data extraction interval covering the entire disturbance transmission process. The window width is initially set based on the theoretical transmission time of the pipeline space span, ensuring complete coverage of the entire process of disturbance generation, transmission, and impact on the sensor.

[0030] like Figure 2 As shown, after the timing alignment window is constructed, the closed-loop control component imports the extracted Doppler frequency shift feature dataset and phase delay difference dataset into the window, performs standardized cross-correlation calculation, and extracts the transient coupling feature vector at the moment of fluid switching by quantifying the correlation between the two sets of timing signals. This feature vector is used as the core input parameter of the fluid dynamics disturbance factor to achieve accurate solution of the disturbance factor. In this embodiment, the cross-correlation calculation is implemented through the following quantization formula: ; Where R(τ) represents the cross-correlation result corresponding to the time offset τ; F d (t) represents the Doppler frequency shift characteristic value at time t; S p (t+τ) represents the phase delay difference at time t after a time offset of τ; T1 is the total duration of the time alignment window. This formula can accurately quantify the temporal correlation between the fluid velocity disturbance signal and the sensor response signal, and eliminate invalid calculation data caused by time misalignment.

[0031] Preferably, in this embodiment, the closed-loop control component can dynamically adjust the width of the timing alignment window based on the peak offset of the cross-correlation operation. The peak value of the cross-correlation operation result corresponds to the maximum associated timing node of the two sets of signals. When the peak offset deviates from the center position of the window, it indicates that the initial window width cannot adapt to the current disturbance transmission timing. The component automatically extends or shrinks the window boundary to ensure that all effective coupled signals are completely acquired. At the same time, the component sets an adaptive sliding step size within the adjusted timing alignment window. The value of the sliding step size is positively correlated with the signal sampling frequency. The sliding step size enables continuous frame-by-frame comparison of signals within the window, extracting the correlation components of Doppler frequency shift features and phase delay difference frame by frame. The feature parameters corresponding to the maximum mutual information are selected as the principal components of the transient coupling feature vector. Meanwhile, redundant and low-correlation interference components in the vector are removed based on a preset information entropy threshold to ensure the purity and accuracy of the disturbance factor input parameters.

[0032] To clearly demonstrate the timing alignment window parameters and cross-correlation operation characteristics under different pipeline spans in this embodiment, the core operating parameters are set in stages according to the pipeline span: when the pipeline span is 0.5–1.0 m, the initial timing window width is 50 ms, the adaptive sliding step size is 50 μs, the maximum cross-correlation threshold is 0.85, and the information entropy redundancy removal threshold is 0.20; when the pipeline span is 1.0–2.0 m, the initial timing window width is 80 ms, the adaptive sliding step size is 60 μs, and the maximum cross-correlation threshold is 0.85. The mutual information threshold is 0.82, and the information entropy redundancy removal threshold is 0.22. When the pipeline space span is 2.0–3.0 m, the initial timing window width is 120 ms, the adaptive sliding step size is 70 μs, the maximum mutual information threshold is 0.80, and the information entropy redundancy removal threshold is 0.25. When the pipeline space span is 3.0–5.0 m, the initial timing window width is 180 ms, the adaptive sliding step size is 80 μs, the maximum mutual information threshold is 0.78, and the information entropy redundancy removal threshold is 0.28.

[0033] The larger the pipeline span, the longer the transmission delay of fluid disturbance from the switching valve to the sensor, so the width of the initial timing window increases synchronously; the sliding step size gradually increases with the increase of pipeline span, which can reduce the computational redundancy of signal comparison in long-distance pipelines; the mutual information threshold and information entropy threshold are adaptively adjusted with changes in operating conditions to adapt to the signal coupling characteristics under different transmission distances and ensure the accuracy of feature vector extraction.

[0034] This embodiment, based on the architecture of the basic embodiment, refines the complete technical logic of signal feature extraction, adaptive correction of timing window, and purification of feature vector. It accurately achieves timing alignment and coupling matching between fluid disturbance signals and sensor response signals, solves the problem of disturbance factor calculation deviation caused by signal timing misalignment and inaccurate feature extraction in traditional technologies, and further improves the accuracy of dynamic compensation of the system.

[0035] In another preferred embodiment, based on the signal extraction and temporal coupling logic of the aforementioned embodiments, the solution mechanism for the fluid dynamics disturbance factor is further optimized. By constructing a nonlinear fluid dynamics state equation, it adapts to the complex flow field changes caused by the viscosity of rainwater and the elastic deformation of the pipe wall, achieving a high-precision quantitative solution for the disturbance factor. Rainwater is different from pure water; during rainfall, it contains silt, suspended solids, and soluble impurities, causing the fluid viscosity coefficient to change in real time. At the same time, long-term fluid scouring causes slight elastic deformation of the pipe wall, changing the effective cross-sectional area of ​​the flow channel and the state of the fluid boundary layer. Traditional linear calculation models cannot adapt to such nonlinear changing characteristics.

[0036] In this embodiment, the closed-loop control component collects the viscosity coefficient and elastic modulus of the current rainwater fluid in real time. The viscosity coefficient is obtained in real time through the fluid viscosity detection unit, characterizing the flow resistance characteristics of the rainwater fluid; the elastic modulus of the pipe wall is collected through the pipe deformation sensing unit, characterizing the elastic deformation capability of the pipe under fluid pressure. The closed-loop control component imports these two types of basic parameters into the nonlinear fluid dynamics state equation to construct a flow field calculation model adapted to real-time fluid conditions. This equation can comprehensively cover the multiple coupled effects of fluid viscous resistance, pipe wall elastic deformation, and sudden changes in fluid velocity.

[0037] The nonlinear fluid dynamics state equation used in this embodiment is as follows: ; Where ρ represents the density of rainwater fluid; μ represents the instantaneous rate of change of fluid velocity; μ represents the real-time fluid viscosity coefficient. The second-order spatial gradient of the flow velocity is represented by E; the elastic modulus of the channel wall is represented by E; and the real-time deformation of the channel wall is represented by ε. F represents the pressure gradient of the fluid in the pipeline. dist This represents the instantaneous disturbance force caused by fluid switching. The equation fully couples multiple physical quantities such as fluid viscosity, pipe wall elastic deformation, dynamic changes in flow velocity, and pressure gradient, and can accurately characterize the nonlinear dynamic characteristics of the rainwater fluid switching process.

[0038] The closed-loop control component substitutes the acquired fluid velocity abrupt change signal and transient response signal into the aforementioned nonlinear fluid dynamics state equation, and calculates two core flow field characteristic parameters—the fluid boundary layer separation index and the eddy current dissipation rate—through an iterative solution method. The fluid boundary layer separation index quantifies the degree to which the fluid boundary layer detaches from the pipe wall. Boundary layer separation directly causes turbulent fluid flow at the sensor detection end face, leading to signal drift. The eddy current dissipation rate quantifies the generation, development, and decay rate of eddies inside the pipe at the moment of switching. The dynamic changes in eddies are the core cause of transient response fluctuations in the sensor.

[0039] Furthermore, the closed-loop control component performs multivariate nonlinear fitting calculations on the fluid boundary layer separation index and eddy current dissipation rate to eliminate the dimensional differences and mutual interference between the two types of parameters, generating a hydrodynamic disturbance factor characterizing the mechanical impulse disturbance at the moment of fluid switching. The nonlinear fitting employs a quadratic polynomial fitting algorithm, achieving accurate fusion of the two types of flow field parameters through weight allocation. The fitting calculation formula is as follows: F=α·S 2 +β·η 2 +γ1·S·η; Where F represents the final solution of the fluid dynamics disturbance factor; S represents the fluid boundary layer separation index; η represents the eddy current dissipation rate; α, β, and γ1 are preset fitting weight coefficients, which correspond to the weight ratios of boundary layer separation, eddy current dissipation, and coupling terms of the two types of parameters, respectively. The coefficient values ​​are determined through a large number of fluid operating conditions calibrations.

[0040] like Figure 3 As shown, preferably, in this embodiment, the closed-loop control component calculates the Reynolds number in the pipeline at the front end of the remote multi-channel switching valve in real time. The Reynolds number is a core parameter for distinguishing between laminar and turbulent flow states, and its calculation formula is as follows: ; Where Re represents the Reynolds number of the pipeline fluid; v represents the average flow velocity of the fluid; d represents the equivalent pipe diameter; and ρ and μ represent the fluid density and viscosity coefficient, respectively. The system presets a critical Reynolds number threshold of 2300 for laminar and turbulent flow. When the real-time calculated Reynolds number crosses this critical threshold, it indicates a sudden change in the flow state of the pipeline fluid, and the closed-loop control component immediately triggers the reconstruction of the boundary conditions of the nonlinear fluid dynamics state equation.

[0041] The specific boundary condition adaptive correction logic is as follows: When the Reynolds number is greater than 2300, the fluid is in a turbulent state, and a large number of disordered eddies are generated inside the pipe. The component introduces an eddy viscosity correction term into the nonlinear fluid dynamics state equation to compensate for the calculation deviation caused by eddy viscosity under turbulent conditions. At the same time, the solution step size of the equation is increased to adapt to the rapidly changing turbulent flow field. When the Reynolds number is less than 2300, the fluid is in a laminar state, and the fluid in the pipe is mainly in an ordered laminar flow. The friction loss of the pipe wall is the main source of disturbance. The component introduces a friction loss correction term into the equation to offset the flow field attenuation deviation caused by pipe wall friction. At the same time, the solution step size of the equation is reduced to improve the calculation accuracy under laminar steady-state conditions.

[0042] To intuitively illustrate the parameter correction and computational characteristics under different flow regimes, the flow regime correction parameters are configured according to the fluid flow regime: for laminar flow, the Reynolds number range is Re < 2300, a friction loss correction term is used, the equation solution step size is 20 μs, and the perturbation factor solution accuracy level is high; for transitional flow, the Reynolds number range is 2300 ≤ Re ≤ 4000, a two-way coupling correction term is used, the equation solution step size is 35 μs, and the perturbation factor solution accuracy level is medium to high; for turbulent flow, the Reynolds number range is Re > 4000, an eddy viscosity correction term is used, the equation solution step size is 50 μs, and the perturbation factor solution accuracy level is dynamic adaptation accuracy.

[0043] In the transitional flow state, two types of correction terms are introduced simultaneously to achieve a smooth transition between laminar and turbulent flow conditions. The solution step size increases synchronously with the degree of flow turbulence, which reduces the consumption of computing resources under high-speed dynamic conditions while ensuring computational accuracy and improving the stability of system operation.

[0044] This embodiment constructs a nonlinear fluid dynamics state equation and combines it with Reynolds number discrimination to achieve adaptive correction of the flow regime. It accurately quantifies the fluid disturbance characteristics under different rainwater conditions, solves the technical problems that traditional linear models cannot adapt to nonlinear fluid changes and have large deviations in disturbance calculation under complex flow regimes, and improves the robustness and adaptability of the disturbance factor solution.

[0045] In a preferred embodiment, based on the disturbance factor solution logic of the aforementioned embodiments, the generation mechanism of dynamic compensation coefficients is further optimized. Through frequency domain transformation and coherence analysis, accurate separation and differentiated compensation of drift components are achieved, generating multi-band dynamic compensation coefficients to adapt to disturbances and drift signals with different frequency characteristics. Traditional time-domain compensation methods cannot distinguish between steady-state baseline drift and instantaneous pulse disturbance drift, and are prone to overcompensation or undercompensation. This embodiment achieves accurate separation and targeted compensation of the two types of drift components through frequency domain spatial operations.

[0046] In this embodiment, the closed-loop control component first uses a Fast Fourier Transform (FFT) algorithm to synchronously transform the time-domain fluid dynamics disturbance factor and the baseline drift of the sensing signal to the frequency domain, obtaining the frequency domain distribution datasets corresponding to the two types of signals. The FFT calculation formula is as follows: ; Where X(k) represents the signal amplitude corresponding to the k-th frequency point in the frequency domain; x(n) represents the original signal value at the n-th sampling point in the time domain; M represents the number of sampling points in a single Fourier transform; and j is the imaginary unit. This formula allows a continuous time-domain signal to be decomposed into frequency-domain components with multiple frequencies and amplitudes, achieving a refined breakdown of signal characteristics.

[0047] like Figure 4 As shown, after completing the frequency domain transformation, the closed-loop control component extracts the frequency domain distribution characteristics of the hydrodynamic disturbance factor and the baseline drift of the sensing signal, respectively. The frequency domain distribution characteristics include core parameters such as the signal's dominant frequency, amplitude spectrum, and power spectral density. The component calculates the coherence index of the two sets of frequency domain signals using a coherence calculation formula. This index is used to quantify the frequency domain correlation between the disturbance factor and the baseline drift. The coherence index calculation formula is as follows: ; Where, γ 2 (f) represents the coherence index at frequency f, with a value ranging from 0 to 1; P FD (f) represents the cross-power spectral density of the perturbation factor and the baseline drift; P F (f) represents the self-power spectral density of the hydrodynamic disturbance factor; P D (f) represents the autopower spectral density of the baseline drift of the sensing signal. The closer the coherence index is to 1, the higher the correlation between the baseline drift at that frequency and the hydrodynamic disturbance.

[0048] Based on the numerical distribution characteristics of the coherence index, the closed-loop control component classifies and separates the frequency domain drift components into steady-state drift components and transient disturbance components. Among them, the low-frequency, high-coherence continuous drift components are steady-state drift components, mainly caused by slow-changing factors such as long-term fluid erosion in pipelines and inherent temperature drift of sensors; the high-frequency, instantaneously abrupt drift components are transient disturbance components, mainly caused by instantaneous dynamic disturbances such as sudden changes in flow velocity and eddy current impacts during fluid switching.

[0049] Furthermore, the closed-loop control component assigns differentiated weighting coefficients to the two types of drift components. Steady-state drift components, with their slow change rate and long influence period, are assigned stable weighting coefficients; transient disturbance components, with their drastic changes and strong instantaneous interference, are assigned high-response weighting coefficients, achieving targeted compensation for different types of drift. After the weighting coefficients are assigned, the component performs an inverse frequency domain transformation using an inverse fast Fourier transform, restoring the frequency domain compensation parameters to the time domain space, generating dynamic compensation coefficients containing multi-band compensation vectors. These multi-band compensation vectors include low-frequency steady-state compensation components, high-frequency transient compensation components, and mid-frequency transition compensation components, enabling accurate compensation for signal drift across the entire frequency band.

[0050] Preferably, in this embodiment, the closed-loop control component calculates the difference between the coherence index and the preset coherence benchmark in real time to obtain the coherence attenuation gradient. The preset coherence benchmark is the steady-state coherence threshold of 0.9 calibrated by the system. The formula for calculating the coherence attenuation gradient is as follows: ; in, γ represents the coherence attenuation gradient; γ0 represents the preset coherence reference value; This represents the average coherence index across the entire frequency band. The component monitors the rate of change of the coherence attenuation gradient in real time and dynamically adjusts the allocation ratio of the differentiated weighting coefficients. When the rate of change of the coherence attenuation gradient exceeds the preset steady-state range, it indicates an increase in the intensity of the current transient fluid disturbance. The system automatically increases the weighting ratio of the transient disturbance component to strengthen the compensation for the transient disturbance. When the rate of change of the attenuation gradient is within the steady-state range, the weighting ratio of the steady-state drift component is maintained to ensure the long-term stability of the baseline.

[0051] The component recalculates the multi-band compensation vector iteratively based on the adjusted weight coefficient ratio, and feeds the updated compensation vector back to the inverse frequency domain conversion input, forming an iterative update closed loop to continuously optimize the adaptability of the dynamic compensation coefficient and avoid compensation lag and compensation failure caused by fixed weights.

[0052] To clearly illustrate the weight allocation rules under different coherence decay gradients, the weight coefficient adaptation parameters are set in stages according to the rate of change of the coherence decay gradient: when the rate of change of the coherence decay gradient is less than 0.01 / s (steady-state interval), the weight ratio of the steady-state drift component is 0.70, the weight ratio of the transient disturbance component is 0.30, and the compensation vector iteration frequency is 10Hz; when the rate of change of the coherence decay gradient is 0.01~0.03 / s (weak fluctuation interval), the weight ratio of the steady-state drift component is 0.55, and the weight ratio of the transient disturbance component is 0.30. The weight ratio is 0.45, and the iteration frequency of the compensation vector is 20Hz. When the coherence decay gradient change rate is 0.03~0.05 / s (strong fluctuation range), the weight ratio of the steady-state drift component is 0.40, the weight ratio of the transient disturbance component is 0.60, and the iteration frequency of the compensation vector is 30Hz. When the coherence decay gradient change rate is greater than 0.05 / s (severe disturbance range), the weight ratio of the steady-state drift component is 0.25, the weight ratio of the transient disturbance component is 0.75, and the iteration frequency of the compensation vector is 50Hz.

[0053] As the rate of change of the coherence decay gradient increases, the system gradually shifts the weight proportion to the transient disturbance component, while increasing the iteration frequency of the compensation vector to quickly adapt to the compensation requirements under severe disturbance conditions, ensuring the real-time performance and accuracy of dynamic compensation.

[0054] like Figure 5 As shown, further, after the closed-loop control component generates the multi-band dynamic compensation coefficients, it performs hierarchical analysis on the multi-band compensation vectors in the compensation coefficients, mapping the compensation vector components of different frequency bands to a non-uniform subdivision drive pulse sequence of the remote multi-channel switching valve stepper motor. Unlike traditional uniformly divided pulse sequences, the non-uniform subdivision drive pulse sequence generated in this embodiment uses pulse step angles arranged in a non-arithmetic sequence. The step angle size adaptively adjusts with the frequency band amplitude of the compensation coefficients. The low-frequency steady-state compensation component corresponds to a small-angle uniform pulse, ensuring stable steady-state operation; the high-frequency transient compensation component corresponds to a large-angle differentiated pulse, quickly responding to instantaneous disturbance compensation requirements.

[0055] Simultaneously, the closed-loop control component dynamically adjusts the duty cycle combination of the non-uniform subdivision drive pulse sequence based on the overall amplitude of the dynamic compensation coefficient. Through the differentiated ratio of multiple pulse duty cycles, it controls the stepper motor to drive the valve core to perform variable acceleration rotation, abandoning the traditional constant angular velocity rotation mode. During switching moments with high disturbance intensity, the pulse duty cycle difference is increased to enhance the speed variation of the valve core rotation and quickly counteract fluid disturbances. Under steady-state fluid conditions, the pulse duty cycle is balanced to maintain stable valve core operation. Simultaneously, based on the timing characteristics of the compensation coefficient, the dwell time node of the switching interval sequence is dynamically modified, adaptively adjusting the steady-state waiting time after the fluid channel is opened, achieving precise matching between the switching action timing and the sensor signal stabilization process.

[0056] This embodiment achieves refined generation of dynamic compensation coefficients and precise control of device drive through the complete logic of frequency domain decomposition, coherence analysis, differential weight iteration, and non-uniform pulse drive. It solves the technical problems of traditional single time domain compensation being unable to adapt to multi-frequency disturbances and having poor adaptability to operating conditions, and greatly improves the measurement stability of the system under complex dynamic fluid conditions.

[0057] In the fourth preferred embodiment, based on the core architecture and operational logic of all the aforementioned embodiments, auxiliary monitoring and anti-interference components are added to further improve the system's multi-dimensional disturbance suppression and error elimination capabilities, adapting to rainwater monitoring scenarios in complex pipe networks and complex environments. This embodiment adds a photoelectric coupling monitoring array and a piezoelectric micro-frequency vibration component arranged along the outlet pipeline of the remote multi-channel switching valve. Simultaneously, a temperature distribution gradient collector is added between the switching valve and the sensor array to respectively achieve microbubble disturbance suppression and thermal gradient interference elimination.

[0058] like Figure 6 As shown, microbubbles are easily generated in the rainwater fluid inside the pipeline during rapid switching and pressure changes. The aggregation of microbubbles changes the actual flow cross-sectional area of ​​the fluid in the pipeline, causing local flow field turbulence. At the same time, bubbles attached to the sensor detection end face can block the sensing area, causing sensor signal drift. Conventional fluid disturbance compensation algorithms cannot eliminate the fixed interference caused by microbubbles. In this embodiment, the photoelectric coupling monitoring array is uniformly arranged along the circumference and axial direction of the outlet pipeline, including multiple sets of light-emitting units and photosensitive acquisition units. The light-emitting units emit a detection light source of fixed wavelength. After the light source penetrates the pipeline fluid, the photosensitive units collect the transmitted light signal to obtain a real-time image of the light transmittance distribution of the fluid in the pipeline.

[0059] In the pure fluid region, the light transmittance is uniform and high. However, in the microbubble aggregation region, light scattering and occlusion occur, forming discrete low transmittance areas. The closed-loop control component accurately identifies these discrete low transmittance areas in the light transmittance distribution image through image grayscale threshold comparison and region contour recognition algorithms, using these areas as microbubble distribution features. The component further calculates the microbubble aggregation density and spatial distribution centroid through pixel statistics and spatial coordinate calibration. The aggregation density represents the number of microbubbles per unit pipe volume, and the spatial distribution centroid represents the axial and radial position coordinates of the microbubble aggregation within the pipe.

[0060] The closed-loop control component generates a matching resonant frequency command based on the microbubble aggregation density and spatial centroid, using the bubble resonant frequency calibration formula. The bubble resonant frequency calculation formula is as follows: ; Among them, f bγ represents the inherent resonant frequency of the microbubble; P0 represents the surface tension coefficient of the rainwater fluid; ρ represents the static pressure of the fluid in the pipeline; and r represents the average radius of the microbubble. The system corrects the amplitude of the resonant frequency based on the microbubble aggregation density. The higher the aggregation density, the larger the corresponding resonant frequency command amplitude, ensuring efficient bubble dissipation.

[0061] The closed-loop control component outputs the generated resonant frequency command to the piezoelectric micro-frequency vibration component, driving the component to generate high-frequency micro-amplitude oscillation deformation at the pipeline position corresponding to the centroid of the microbubble spatial distribution. The high-frequency vibration can disrupt the aggregation and attachment state of the microbubbles, accelerate the rupture and dissipation of the microbubbles, and eliminate the flow field turbulence and sensor signal interference caused by the aggregation of microbubbles.

[0062] Meanwhile, uneven axial temperature distribution in the pipeline can cause localized variations in fluid viscosity and density, resulting in thermally induced baseline drift. This drift component couples with fluid dynamic disturbance drift, interfering with the accuracy of convolution compensation calculations. In this embodiment, a temperature distribution gradient acquisition unit is deployed with multiple temperature detection units along the pipeline axis to simultaneously acquire real-time temperature sequences at multiple axial points. A thermal gradient field vector is constructed along the fluid flow direction using a spatial interpolation algorithm. This thermal gradient field vector can accurately characterize the spatial temperature distribution differences of the fluid in the pipeline.

[0063] The closed-loop control component performs orthogonal decoupling operations on the thermal gradient field vector and the hydrodynamic disturbance factor. The orthogonal decoupling operation formula is as follows: ; Among them, F pure The formula represents the decoupled pure hydrodynamic disturbance factor; F represents the original hydrodynamic disturbance factor; and T represents the thermal gradient field vector. This formula allows for the complete separation of the thermally induced baseline drift component affected by the thermal gradient field vector, accurately distinguishing between hydrodynamic disturbance drift and temperature disturbance drift.

[0064] The closed-loop control component subtracts the thermally induced baseline drift component from the original sensor signal baseline drift, and performs a convolution operation between the remaining drift component caused by pure fluid disturbance and the corrected fluid dynamic disturbance factor to generate a more accurate dynamic compensation coefficient, thereby eliminating computational interference caused by temperature gradient.

[0065] This embodiment supplements the environmental and media interference factors not covered by traditional fluid disturbance compensation by adding a microbubble monitoring and suppression structure and a thermal gradient decoupling structure. It achieves multi-dimensional comprehensive suppression of mechanical disturbance, bubble disturbance and temperature gradient disturbance, and further improves the measurement stability and calculation accuracy of the system in complex field environments.

Claims

1. A remote fluid switching type rainwater multi-parameter dynamic online analysis system, characterized in that, This includes a remote multi-channel switching valve, a multi-parameter sensor array, and closed-loop control components; The closed-loop control component collects the fluid velocity change signal at the moment of switching of the remote multi-channel switching valve and the transient response signal of the multi-parameter sensor array in real time, and calculates the fluid dynamic disturbance factor at the moment of switching. The fluid dynamics disturbance factor is convolved with the baseline drift of the sensing signal of the multi-parameter sensor array to generate dynamic compensation coefficients. The closed-loop control component adjusts the step drive pulse width and switching interval timing of the remote multi-channel switching valve in reverse based on the dynamic compensation coefficient, so that the fluid switching action and the sensor signal stabilization process are deeply coupled.

2. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to claim 1, characterized in that, The closed-loop control component extracts the Doppler frequency shift characteristics of the pressure fluctuation inside the flow channel of the remote multi-channel switching valve body as the fluid velocity change signal, and simultaneously extracts the phase delay difference of the output electrical signal of the multi-parameter sensor array as the transient response signal. The closed-loop control component constructs a timing alignment window based on the pipeline space span between the remote multi-channel switching valve and the multi-parameter sensor array. It performs cross-correlation calculation on the Doppler frequency shift feature and the phase delay difference within the timing alignment window, and extracts the transient coupling feature vector at the moment of fluid switching as the input parameter of the fluid dynamics disturbance factor.

3. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to claim 2, characterized in that, The closed-loop control component dynamically adjusts the width of the timing alignment window based on the peak offset of the cross-correlation calculation, and establishes a sliding step size within the timing alignment window; The Doppler frequency shift feature and the phase delay difference are compared frame by frame within the time alignment window according to the sliding step size. The maximum mutual information between the Doppler frequency shift feature and the phase delay difference is extracted as the principal component of the transient coupling feature vector. Redundant components in the transient coupling feature vector that are below the information entropy threshold are removed.

4. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to any one of claims 1 to 3, characterized in that, The closed-loop control component constructs a nonlinear fluid dynamics state equation based on the viscosity coefficient of the current rainwater fluid and the elastic modulus of the channel wall. The fluid velocity change signal and the transient response signal are input into the nonlinear fluid dynamics state equation to solve for the fluid boundary layer separation index and eddy current dissipation rate. The closed-loop control component performs nonlinear fitting between the fluid boundary layer separation index and the eddy current dissipation rate to generate the fluid dynamic disturbance factor characterizing the mechanical impulse disturbance at the moment of fluid switching.

5. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to claim 4, characterized in that, The closed-loop control component calculates the Reynolds number in the pipeline at the front end of the remote multi-channel switching valve in real time. When the Reynolds number crosses the critical threshold between laminar and turbulent flow, it triggers the boundary condition reconstruction of the nonlinear fluid dynamics state equation. The closed-loop control component introduces an eddy viscosity correction term under turbulent boundary conditions and a friction loss correction term under laminar boundary conditions. Based on the eddy viscosity correction term or the friction loss correction term, it adaptively switches the solution step size of the nonlinear fluid dynamics state equation and outputs the corrected fluid dynamics disturbance factor.

6. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to any one of claims 1 to 3, characterized in that, The closed-loop control component converts the hydrodynamic disturbance factor and the baseline drift of the sensing signal to the frequency domain space, and calculates the coherence index between the frequency domain distribution characteristics of the hydrodynamic disturbance factor and the frequency domain distribution characteristics of the baseline drift of the sensing signal in the frequency domain space. Based on the coherence index, steady-state drift components and transient disturbance components are separated. Differentiated weighting coefficients are assigned to the steady-state drift components and the transient disturbance components, and inverse frequency domain transformation is performed to generate the dynamic compensation coefficients containing multi-band compensation vectors.

7. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to claim 6, characterized in that, The closed-loop control component performs a difference calculation between the coherence index and a preset coherence benchmark to obtain the coherence attenuation gradient. The allocation ratio of the differentiated weight coefficients is dynamically adjusted according to the rate of change of the coherence decay gradient. When the rate of change of the coherence decay gradient exceeds the steady-state range, the weight coefficient ratio of the transient disturbance component is increased. The multi-band compensation vector is regenerated based on the adjusted weight coefficient ratio, and the multi-band compensation vector is fed back to the input of the inverse frequency domain conversion for iterative update.

8. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to any one of claims 1 to 3, characterized in that, The closed-loop control component parses the multi-band compensation vector in the dynamic compensation coefficient and maps the multi-band compensation vector to the non-uniform subdivision drive pulse sequence of the stepper motor in the remote multi-channel switching valve. The non-uniform subdivision driving pulse sequence includes pulse step angles arranged in a non-arithmetic sequence; The closed-loop control component adjusts the duty cycle combination of the non-uniform subdivision drive pulse sequence according to the amplitude of the dynamic compensation coefficient, so as to control the valve core of the remote multi-channel switching valve to perform variable acceleration rotation and dynamically modify the dwell time node of the switching interval sequence.

9. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to any one of claims 1 to 3, characterized in that, It also includes a photoelectric coupling monitoring array and a piezoelectric micro-frequency vibration component arranged along the outlet pipeline of the remote multi-channel switching valve; The optocoupler monitoring array acquires an image of the light transmittance distribution of the fluid flowing through the pipeline, and identifies discrete low transmittance regions in the light transmittance distribution image as microbubble distribution features. The closed-loop control component calculates the microbubble aggregation density and spatial distribution centroid based on the microbubble distribution characteristics, generates a resonant frequency command based on the microbubble aggregation density and the spatial distribution centroid, and drives the piezoelectric micro-frequency vibration component to generate a high-frequency micro-amplitude oscillation deformation at the pipeline position corresponding to the spatial distribution centroid that matches the resonant frequency command.

10. The remote fluid switching type rainwater multi-parameter dynamic online analysis system according to any one of claims 1 to 3, characterized in that, It also includes a temperature distribution gradient collector arranged between the remote multi-channel switching valve and the multi-parameter sensor array; The temperature distribution gradient collector acquires the temperature sequence at multiple points along the axial direction of the pipeline and constructs a thermal gradient field vector along the fluid flow direction. The closed-loop control component performs orthogonal decoupling operation on the thermal gradient field vector and the hydrodynamic disturbance factor, separates the thermally induced baseline drift component affected by the thermal gradient field vector, subtracts the thermally induced baseline drift component from the baseline drift of the sensing signal, and performs the convolution operation on the remaining drift component and the hydrodynamic disturbance factor.

Images