Water affair pump station intelligent energy-saving scheduling system based on digital twinning

Abstract

The present application relates to the technical field of digital twinning and intelligent water control, and particularly relates to a water pump station intelligent energy-saving scheduling system based on digital twinning, which comprises: an ultra-transient heterogeneous tensor high-frequency sensing module that collects motor and fluid data and outputs electromagnetic and fluid transient state vectors; a symplectic geometry electromechanical hydraulic cross-coupling twin module that calculates electromechanical hydraulic cross-interference tensors, constructs a system cross-coupling Hamilton energy function in combination with anti-cavitation differential constraint rules; a non-complete constraint symplectic optimization control module that uses a symplectic integral algorithm for iterative solution and outputs an optimal control matrix; a space vector inverter physical execution module that analyzes the matrix to drive a variable frequency inverter device to execute physical variable frequency; and a high-order harmonic residual dynamic closed loop module that obtains a high-order harmonic residual vector generated by execution and feeds back to correct the energy function. The present application solves the problem of electromechanical hydraulic resonance induced by the split of electromagnetic transient and fluid transient under ultra-high frequency conversion, and realizes precise closed loop of full-stack logic and bottom-layer stability preservation.

Description

Technical Field

[0001] This invention relates to the field of digital twin and smart water control technology, and in particular to a smart energy-saving scheduling system for water pumping stations based on digital twins. Background Technology

[0002] Large-scale cascade water pumping stations are the physical hubs of municipal water supply and inter-basin water transfer projects. In recent years, digital twin technology has been gradually applied to the variable frequency energy-saving scheduling of pumping station units. Most existing water digital twin scheduling systems remain at the level of macroscopic three-dimensional visualization or steady-state hydraulic balance. In terms of algorithmic control logic, existing technologies typically treat the pump's drive source as an ideal torque output black box, relying solely on open-loop optimization calculations based on the network's water demand and pipe resistance models, and then directly issuing linear frequency adjustment commands to the inverter. This conventional scheduling architecture fundamentally severs the cross-physical interference between the electromagnetic transient evolution process inside the motor and the transient damping process of the external network's fluid dynamics.

[0003] When a cascaded pipeline network encounters sudden topological changes such as a burst pipe or an extreme water surge, forcing the dispatching system to instruct the underlying units to perform ultra-transient, high-speed frequency conversion, the rapid change in the inverter's output voltage vector can lead to saturation of the motor's stator and rotor flux linkages, resulting in significant high-frequency electromagnetic harmonic losses. Simultaneously, this instantaneous change in nonlinear electromagnetic torque directly induces nonlinear water hammer pressure waves within the pipeline network manifold space through the pump impeller. Because existing dispatching models fail to couple these multi-physics fields, the high-frequency electromagnetic transient oscillations and the fluid transient pressure waves are highly prone to physical frequency overlap, inducing highly destructive transient electromechanical-hydraulic resonances. This single core technical problem directly causes the pumping station units to easily trigger motor overcurrent trip protection during the transient state of ultra-speed frequency adjustment, and induces microscopic cavitation spalling of the pump impeller, compromising the operational stability of the underlying hardware and generating extreme transient power dissipation. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides an intelligent energy-saving scheduling system for water pumping stations based on digital twins. It aims to improve upon the traditional pumping station scheduling system, which mostly adopts independent models that separate electromagnetic and fluid components. Due to the neglect of the frequency overlap of multiple physical fields under extremely high frequency conversion, transient electromechanical-hydraulic resonance is easily induced.

[0005] This invention provides the following technical solution: an intelligent energy-saving scheduling system for water pumping stations based on digital twins, comprising the following modules:

[0006] The ultra-transient heterogeneous tensor high-frequency sensing module is used to collect motor data and fluid data of the water pumping station and output the electromagnetic transient state vector of the motor and the transient state vector of the fluid.

[0007] The symplectic geometric electromechanical-hydraulic cross-coupling twin module is used to receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid, calculate the electro-hydraulic cross-interference tensor, and construct the system cross-coupling Hamiltonian energy function in combination with the preset anti-cavitation differential constraint rules.

[0008] The nonholonomic constraint symplectic optimization control module is used to iteratively solve the cross-coupled Hamiltonian energy function of the system using a symplectic integral algorithm and output the optimal control matrix.

[0009] The space vector inverter physical execution module is used to analyze the optimal control matrix to drive the variable frequency inverter device of the water pumping station and perform physical variable frequency control actions.

[0010] The high-order harmonic residual dynamic closed-loop module is used to obtain the high-order harmonic residual vector generated when the physical frequency conversion control action is executed, and feed the high-order harmonic residual vector back to the symplectic geometric electromechanical-hydraulic cross-coupled twin module to correct the cross-coupled Hamiltonian energy function of the system.

[0011] By adopting the above technical solution, the Hamiltonian energy function is constructed based on the electromechanical-hydraulic interference tensor and the symplectic closed-loop optimization is performed, thereby realizing cross-physical field collaborative control. This improves the problem that traditional pump station scheduling mostly adopts independent models that separate electromagnetic and fluid, which easily induce transient electromechanical-hydraulic resonance due to neglecting the frequency overlap of multiple physical fields under ultra-fast frequency conversion.

[0012] Optionally, the ultratransient heterogeneous tensor high-frequency sensing module is used to perform:

[0013] Collect the stator transient current data and rotor transient flux data of the frequency converter device of the water pumping station in the three-phase static coordinate system;

[0014] A coordinate transformation operation is performed on the stator transient current data and the rotor transient flux data to map the stationary coordinate system data to the synchronous rotating coordinate system;

[0015] The direct-axis transient current component, the quadrature-axis transient current component, the direct-axis transient flux component, and the quadrature-axis transient flux component are extracted respectively in the synchronous rotating coordinate system.

[0016] The direct-axis transient current component, the quadrature-axis transient current component, the direct-axis transient flux linkage component, and the quadrature-axis transient flux linkage component are tensor-concatenated to generate the electromagnetic transient state vector of the motor.

[0017] Optionally, the ultratransient heterogeneous tensor high-frequency sensing module is further configured to perform:

[0018] Simultaneously collect transient outlet flow data of multiple water pumps at the water pumping station and transient node water pressure fluctuation data of key topology nodes in the wide area pipeline network;

[0019] Extract the first-order time derivative of the transient outlet flow rate data of the water pump and the transient node water pressure fluctuation data within the rolling time window;

[0020] The transient outlet flow rate data of the water pump, the transient node water pressure fluctuation data, and the first-order time derivative are reconstructed into tensors according to a preset dimension to generate the transient state vector of the fluid.

[0021] Optionally, the symplectic geometric electromechanical-hydraulic cross-coupled twin module is used to perform:

[0022] Receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid;

[0023] Establish a symplectic manifold space for mapping the energy evolution characteristics of the system, and extract a Riemannian metric tensor that matches the transient state vector of the fluid and characterizes the transient hydraulic damping characteristics of the pipe network in the symplectic manifold space;

[0024] Extract the initial electromechanical interference matrix characterizing the physical coupling relationship between the electromagnetic torque of the motor and the fluid reaction torque of the pump impeller;

[0025] The Riemannian metric tensor is injected as a nonlinear damping feature into the initial electromechanical interference matrix, and the electromechanical-hydraulic cross-interference tensor is generated by mapping.

[0026] Optionally, the symplectic geometric electromechanical-hydraulic cross-coupled twin module is further configured to perform:

[0027] Extract the anti-cavitation differential constraint rules that reflect the transient rate of fluid state vector triggering local cavitation boundaries;

[0028] Based on the law of conservation of energy, an electromagnetic inertia penalty term is established using the electromagnetic transient state vector of the motor, and a hydraulic damping penalty term is established using the fluid transient state vector and the Riemann metric tensor.

[0029] The electromagnetic inertial penalty term, the hydraulic damping penalty term, the electromechanical-hydraulic cross-interference tensor, and the anti-cavitation differential constraint rule are functionally aggregated in the symplectic manifold space to generate the system cross-coupled Hamiltonian energy function.

[0030] Optionally, the nonholonomic constraint symplectic optimization control module is used to execute:

[0031] Receive the cross-coupled Hamiltonian energy function of the system;

[0032] Establish a symplectic space for system integration, and integrate and map the electromagnetic transient state vector of the motor and the transient state vector of the fluid into a generalized coordinate vector and a generalized momentum vector in the symplectic space for system integration.

[0033] Using the generalized coordinate vector and the generalized momentum vector, a set of continuous-time-domain Hamiltonian canonical equations corresponding to the cross-coupled Hamiltonian energy function of the system is established.

[0034] Optionally, the nonholonomic constraint symplectic optimization control module is further configured to perform:

[0035] Configure the Bausch integral algorithm as an implicit Singlonga-Kutta integrator;

[0036] The implicit Singlonga-Kutta integrator is used to perform high-order discretization of the continuous-time domain Hamiltonian canonical equations to generate an internal nonlinear node constraint equation system.

[0037] Under the constraint boundary of maintaining the conservation of the symplectic geometric area element, the Newton iteration method is used to iteratively solve the constraint equations of the internal nonlinear nodes to extract the stator voltage vector sequence and the frequency conversion rate sequence.

[0038] The optimal control matrix is ​​generated by combining the stator voltage vector sequence with the frequency conversion rate sequence.

[0039] Optionally, the space vector inverter physics execution module is used to execute:

[0040] The optimal control matrix is ​​analyzed and decoded into a space vector pulse width modulation signal sequence;

[0041] The space vector pulse width modulation signal sequence is injected into the underlying drive channel of the frequency inverter device;

[0042] The on-time and off-time of the power switching elements in the frequency converter are controlled according to the space vector pulse width modulation signal sequence, and the space vector pulse width modulation signal sequence is converted into actual electromagnetic torque to perform the physical frequency conversion control action.

[0043] Optionally, the higher harmonic residual dynamic closed-loop module is used to perform:

[0044] During the execution of the physical frequency conversion control action, the physical stator high-order harmonic current components actually output by the frequency conversion inverter device to the motor are simultaneously collected;

[0045] Extract the ideal reference high-order harmonic current components generated by the evolution prediction of the symplectic geometric electromechanical-hydraulic cross-coupled twin module under the same discrete time step;

[0046] The high-order harmonic current components of the physical stator and the high-order harmonic current components of the ideal reference are compared using high-frequency differential calculation to generate the high-order harmonic residual vector.

[0047] Optionally, the higher harmonic residual dynamic closed-loop module is further configured to perform:

[0048] The higher harmonic residual vector is transmitted in reverse to the symplectic geometric electromechanical-hydraulic cross-coupled twin module;

[0049] A weighted compensation correction operator is constructed based on the aforementioned higher harmonic residual vector;

[0050] The weight compensation correction operator is applied to the electromechanical-hydraulic cross-interference tensor to update the coupling weight coefficients inside the electromechanical-hydraulic cross-interference tensor;

[0051] The system cross-coupled Hamiltonian energy function for the next control cycle is reconstructed using the electromechanical-hydraulic cross-interference tensor after updating the coupling weight coefficients.

[0052] The present invention has the following beneficial effects:

[0053] 1. In this invention, Hamiltonian energy function is constructed based on electromechanical-hydraulic interference tensor and optimization is performed using symplectic closed-loop method to achieve cross-physical field collaborative control. This improves the problem that traditional pump station scheduling often adopts independent models that separate electromagnetic and fluid, which easily induce transient electromechanical-hydraulic resonance due to neglecting the frequency overlap of multiple physical fields under ultra-fast frequency conversion.

[0054] 2. In this invention, by extracting the first-order time derivatives of water pressure fluctuations and pipeline flow and reconstructing transient vectors in combination with electromagnetic features, the underlying alternating high-frequency features are captured. This improves the problem that traditional sensing architectures mostly use static steady-state scalars for direct acquisition, which cannot accurately identify millisecond-level fluid abrupt change boundaries, resulting in a serious lag in the response of the control model.

[0055] 3. In this invention, by configuring an implicit Singlonga-Kutta integrator to iteratively solve the problem while maintaining the conserved boundary of the symplectic area element, numerical dissipation in the calculation process is eliminated. This improves the problem that traditional Hamiltonian equations mostly use explicit integration algorithms, where truncation errors are easily accumulated and amplified in nonlinear calculations, resulting in severe distortion of the extremum optimization instructions.

[0056] 4. In this invention, the weights of the electromechanical-hydraulic cross-interference tensor are updated online by extracting the actual and theoretical high-order harmonic residuals, thereby achieving adaptive calibration of the underlying physical boundary. This improves the problem that traditional digital twins mostly use fixed internal parameter configurations, and the mismatch between the physical and virtual mapping caused by the mechanical aging of the equipment leads to the deterioration of the control accuracy during long-term operation. Attached Figure Description

[0057] Figure 1 This is an architecture diagram of an intelligent energy-saving scheduling system for water pumping stations based on digital twins proposed in this invention;

[0058] Figure 2 This is a flowchart of the nonholonomic constraint-preserving symplectic optimization control operation of an intelligent energy-saving scheduling system for water pumping stations based on digital twins, as proposed in this invention. Detailed Implementation

[0059] The technical solutions in 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.

[0060] Example 1:

[0061] In a first embodiment of the present invention, the present invention provides an intelligent energy-saving scheduling system for water pumping stations based on digital twins, such as... Figure 1 As shown, it includes the following steps:

[0062] The ultra-transient heterogeneous tensor high-frequency sensing module is used to collect motor data and fluid data from water pumping stations and output the electromagnetic transient state vector of the motor and the transient state vector of the fluid.

[0063] Furthermore, the ultratransient heterogeneous tensor high-frequency sensing module is used to perform:

[0064] Collect stator transient current data and rotor transient flux data of frequency converter devices in water pumping stations in a three-phase stationary coordinate system;

[0065] A coordinate transformation operation is performed on the stator transient current data and the rotor transient flux data to map the stationary coordinate system data to the synchronous rotating coordinate system;

[0066] The direct-axis transient current component, quadrature-axis transient current component, direct-axis transient flux linkage component, and quadrature-axis transient flux linkage component are extracted respectively in a synchronous rotating coordinate system.

[0067] The direct-axis transient current component, quadrature-axis transient current component, direct-axis transient flux linkage component, and quadrature-axis transient flux linkage component are tensor-concatenated to generate the electromagnetic transient state vector of the motor.

[0068] The ultratransient heterogeneous tensor high-frequency sensing module is also used to perform:

[0069] Simultaneously collect transient outlet flow data from multiple water pumps at water pumping stations, as well as transient node water pressure fluctuation data from key topological nodes in the wide-area pipe network;

[0070] Extract the first time derivative of the transient outlet flow rate data and transient node water pressure fluctuation data of the water pump within the rolling time window;

[0071] The transient outlet flow rate data of the water pump, the transient node water pressure fluctuation data, and the first-order time derivative are reconstructed into a tensor according to a preset dimension to generate a transient fluid state vector.

[0072] Specifically, the ultra-transient heterogeneous tensor high-frequency sensing module, as the underlying physical data interface and dimensionality reduction processing hub of the digital twin system, plays a core role in unifying the dimensions of heterogeneous data from multiple physical fields (electromechanical and hydraulic). The data input of the ultra-transient heterogeneous tensor high-frequency sensing module connects to the electromagnetic sensors at the bottom layer of the water pumping station and the physical sensors at the pipeline nodes. After the original alternating physical signals enter the ultra-transient heterogeneous tensor high-frequency sensing module, they undergo orthogonal transformation of the spatial coordinate system and time-dimension differentiation, outputting dimensionality-reduced and decoupled transient state vectors of the motor electromagnetic system and the fluid transient state vector.

[0073] In the specific technical details of achieving dimensionality reduction of motor data states, it is necessary to use the Parker transformation matrix to map the high-frequency alternating data in the three-phase stationary coordinate system to the synchronous rotating coordinate system, and establish the following mathematical analytical model:

[0074] ;

[0075] ;

[0076] ;

[0077] in , , This represents the stator transient current data in a three-phase stationary coordinate system. This indicates the rotor magnetic pole position angle of the motor, with a value range of [value missing]. to radian. This represents the direct-axis transient current component. This represents the cross-axis transient current component. This represents the direct-axis transient flux linkage component. This represents the cross-axis transient flux linkage component. This represents the electromagnetic transient state vector of the motor.

[0078] In the technical details of realizing fluid data state extraction, the following first-order derivative mathematical model is established based on the transient and abrupt change characteristics of physical fluid dynamics:

[0079] ;

[0080] ;

[0081] ;

[0082] in Indicates the first The water pump at the current sampling time Transient outlet flow rate data of the water pump. Indicates the first At the current sampling time, key topology nodes of the pipeline network Transient node water pressure fluctuation data. This indicates the time step of the rolling time window and should be set to a value in microseconds. Indicates the calculated output of the first... The first time derivative of the flow rate of the water pump. Indicates the calculated output of the first... The first time derivative of water pressure fluctuations at each node. This represents the transient state vector of the fluid.

[0083] By performing coordinate transformation on the stator transient current data and the rotor transient flux linkage data, and extracting the first-order time derivative of the pump transient outlet flow rate data and transient node water pressure fluctuation data within the rolling time window, this improves the traditional technical defect of relying solely on static scalar monitoring, which makes it impossible to identify alternating electromagnetic harmonics and fluid transient rapid change boundaries. This avoids the specific problems of subsequent twin optimization models falling into local deadlock or causing control delay due to the lack of high-frequency physical characteristics.

[0084] The symplectic geometric electromechanical-hydraulic cross-coupling twin module is used to receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid, calculate the electro-hydraulic cross-interference tensor, and construct the system cross-coupling Hamiltonian energy function by combining the preset anti-cavitation differential constraint rules.

[0085] Furthermore, the symmetric geometric electromechanical-hydraulic cross-coupled twin module is used to perform:

[0086] Receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid;

[0087] Establish a symplectic manifold space for mapping the energy evolution characteristics of the system, and extract a Riemannian metric tensor that matches the transient state vector of the fluid and characterizes the transient hydraulic damping characteristics of the pipe network from the symplectic manifold space.

[0088] Extract the initial electromechanical interference matrix characterizing the physical coupling relationship between the electromagnetic torque of the motor and the fluid reaction torque of the pump impeller;

[0089] The Riemannian metric tensor is injected as a nonlinear damping feature into the initial electromechanical interference matrix, and then mapped to generate the electromechanical-hydraulic cross-interference tensor.

[0090] The symmetric geometric electromechanical-hydraulic cross-coupled twin module is also used to perform:

[0091] Extract anti-cavitation differential constraint rules that reflect the transient rate of fluid state vector triggering local cavitation boundary;

[0092] Based on the law of conservation of energy, an electromagnetic inertia penalty term is established using the electromagnetic transient state vector of the motor, and a hydraulic damping penalty term is established using the fluid transient state vector and the Riemannian metric tensor.

[0093] The electromagnetic inertial penalty term, hydraulic damping penalty term, electromechanical-hydraulic cross-interference tensor, and anti-cavitation differential constraint rule are functionally aggregated in the symplectic manifold space to generate the system cross-coupled Hamiltonian energy function.

[0094] Specifically, the symplectic geometric electromechanical-hydraulic cross-coupling twin module serves as the core mapping hub for transforming the system's underlying physical mechanisms into a high-dimensional mathematical optimization space. The module's input receives the electromagnetic transient state vectors of the motor and the transient state vectors of the fluid from the preceding module, while its output generates the system's cross-coupling Hamiltonian energy function for subsequent symplectic integral algorithms.

[0095] The symplectic geometric electromechanical-hydraulic cross-coupling twin module establishes a symplectic manifold space mapping the electromechanical-hydraulic joint energy evolution characteristics of the system. For the received fluid transient state vector, the system establishes a Riemannian metric tensor characterizing the transient hydraulic damping characteristics of the pipe network based on Darcy-Weisbach's friction law. Its specific mathematical expression is configured as follows:

[0096] ;

[0097] in This represents the Riemannian metric tensor extracted from the symplectic manifold space. This represents a mathematical operator for constructing a diagonal matrix. Indicates the first The hydraulic friction coefficient of the branch of the pipe network topology. Indicates the first The equivalent physical length of a branch of a pipeline network topology. Indicates the first The inner diameter of the pipes in the branch of the pipeline network topology. Indicates the first The cross-sectional area of ​​the pipes in the branch of the pipeline network topology. Indicates the first The absolute value of the transient outlet flow rate of the water pump.

[0098] To accurately quantify the cross-physical field interference effect within the pump unit, the system extracts an initial electromechanical interference matrix characterizing the physical coupling relationship between the motor's electromagnetic torque and the pump impeller's fluid reaction torque, based on the electromechanical energy conversion theorem:

[0099] ;

[0100] in This represents the extracted initial electromechanical interference matrix. This indicates the number of pole pairs of the motor in the water pumping station. This indicates the physical density of the fluid being pumped. This represents the gravitational acceleration constant. This indicates the transient volumetric efficiency of the water pump unit.

[0101] Using the above computational model, the following spatial mapping relationship is established to generate the electromechanical-hydraulic cross-interference tensor:

[0102] ;

[0103] in This represents the electromechanical-hydraulic cross-interference tensor generated by the mapping.

[0104] After completing the extraction of cross-interference features, the system invokes the anti-cavitation differential constraint rule that reflects the transient rate of fluid state vector triggering the local cavitation boundary. For the nodal water pressure drop caused by water hammer, a nonlinear penalty barrier function is set as the core calculation formula for the anti-cavitation differential constraint rule:

[0105] ;

[0106] in This represents the constructed anti-cavitation differential constraint rule function. This represents the preset extreme value penalty factor, which is a constant greater than one. It represents the absolute time transient rate of water pressure fluctuation data at nodes in the fluid transient state vector. This represents the critical water pressure transient drop threshold at which micro-cavitation occurs in the pump impeller, and is calibrated based on the effective net positive suction head (NPSH) boundary of the pump.

[0107] Based on the law of conservation of energy, penalty terms for each physical field are constructed and functional aggregation is performed using Lagrange multipliers to establish an analytical model of the system's cross-coupled Hamiltonian energy function:

[0108] ;

[0109] in This represents the system cross-coupled Hamiltonian energy function generated by functional aggregation. This represents the received electromagnetic transient state vector of the motor. This represents the transient inductance parameter matrix of the water pump station motor. The values ​​are determined by the inductance value on the motor's nameplate and the real-time magnetic circuit saturation online calibration. This represents the electromagnetic inertial penalty term established using the electromagnetic transient state vector of the motor. This represents the hydraulic damping penalty term established using the fluid transient state vector and the Riemannian metric tensor. This represents the coupled energy term characterizing the nonlinear cross-work process of electromechanical-hydraulic processes. It represents the first derivative of the fluid transient state vector with respect to time. This represents the differential constraint rule function for cavitation prevention. This represents the Lagrange multiplier vector that introduces cavitation constraints into the energy manifold space.

[0110] By injecting the Riemannian metric tensor as a nonlinear damping feature into the initial electromechanical interference matrix, and functionally aggregating electromagnetic inertia, hydraulic damping, and electromechanical-hydraulic cross-interference in symplectic manifold space, this approach overcomes the technical limitation of traditional scheduling systems that calculate the electromagnetic model of the motor and the hydraulic model of the pipeline separately. This technical solution addresses the technical defect of system instability induced by the frequency overlap of electromagnetic torque pulsation and nonlinear water hammer pressure waves under extreme frequency regulation conditions. It constructs a mathematical boundary containing complete physical interference features for subsequent searches for the global minimum dissipated energy, avoiding the divergence of control commands caused by the lack of cross-physical field interference factors in the optimization objective function.

[0111] like Figure 2 As shown, the nonholonomic constraint symplectic optimization control module is used to iteratively solve the cross-coupled Hamiltonian energy function of the system through the symplectic integral algorithm and output the optimal control matrix.

[0112] Furthermore, the nonholonomic constraint symplectic optimization control module is used to execute:

[0113] The receiving system cross-couples the Hamiltonian energy function;

[0114] Establish a symplectic space for system integration, and integrate and map the electromagnetic transient state vector of the motor and the transient state vector of the fluid into a generalized coordinate vector and a generalized momentum vector in the symplectic space for system integration.

[0115] By using generalized coordinate vectors and generalized momentum vectors, a set of continuous-time-domain Hamiltonian canonical equations corresponding to the cross-coupled Hamiltonian energy function of the system is established.

[0116] The nonholonomic constraint symplectic optimization control module is also used to execute:

[0117] Configure the Baoxin integral algorithm as an implicit Singlonga-Kutta integrator;

[0118] The continuous-time domain Hamiltonian canonical equations are advanced by using an implicit Singlonga-Kutta integrator to perform high-order discretization processing, thereby generating an internal nonlinear node constraint equation system.

[0119] Under the constraint boundary of maintaining the conservation of symplectic geometric area element, the Newton iteration method is used to iteratively solve the internal nonlinear node constraint equation system to extract the stator voltage vector sequence and frequency conversion rate sequence.

[0120] The stator voltage vector sequence is combined with the frequency conversion speed sequence to generate the optimal control matrix.

[0121] Specifically, the nonholonomic constraint-preserving symplectic optimization control module serves as the central solution hub of the entire digital twin system. Its data input receives the system cross-coupled Hamiltonian energy function generated by the preceding mapping, and performs numerical optimization from the continuous-time domain to the discrete-time domain within the internally established system integral symplectic space. Its data output generates the optimal control matrix, which is directly distributed to the underlying device's physical drive.

[0122] The system integrates and maps the electromagnetic transient state vector of the motor and the transient state vector of the fluid into a generalized coordinate vector and a generalized momentum vector, and uses the generalized coordinate vector and generalized momentum vector to establish a continuous-time domain Hamiltonian canonical equation system:

[0123] ;

[0124] in This represents the generalized coordinate vector generated by integrating the mapping in the symplectic space of the system integral operation. This represents the generalized momentum vector generated by the integration mapping in the symplectic space of the system integral operation. These represent the first derivatives of the generalized coordinate vector and the generalized momentum vector with respect to time, respectively. Represents a standard symplectic matrix, with values ​​ranging from 1 to 2. ,in It is an identity matrix. This represents the Jacobian gradient vector of the system's cross-coupled Hamiltonian energy function with respect to the variables. This represents the control input coupling matrix. This represents the control input vector to be solved.

[0125] To solve the equations, the Bausch-Stokes integral algorithm is configured as an implicit Synunge-Kutta integrator. To accommodate the real-time computing power constraints of the water pumping station's underlying controller, the implicit Synunge-Kutta integrator is specifically configured as a second-order Gauss-Legendé implicit symplectic scheme, with a set number of propagation stages. Configure Runge-Kutta quadrature weight coefficients that satisfy the symplectic orthogonality condition. The implicit Singlonga-Kutta integrator performs high-order discretization processing on the continuous-time domain Hamiltonian canonical equations, generating an internal nonlinear nodal constraint equation system:

[0126] ;

[0127] ;

[0128] in , They represent the first time. Discrete time step and the first The generalized coordinate vector of the discrete time step. , They represent the first time. Discrete time step and the first The generalized momentum vector at the discrete time step. This represents the discrete time step, with values ​​in the microsecond range. This represents the number of propulsion stages in the implicit Singlonga-Kutta integrator. Let represent the Runge-Kutta integrand weight coefficients that satisfy the symplectic orthogonality condition. , They represent the first time. The coordinate slope vector and momentum slope vector at each intermediate computation node.

[0129] Under strict boundary constraints that maintain the conservation of symplectic geometric area elements, the system employs Newton's iteration method to iteratively solve the internal nonlinear nodal constraint equations. To prevent computational deadlock at the lower levels, the system extracts the first... The known generalized coordinates and generalized momentum of the discrete time step are used as the first... The initial value vector for the Newton iteration is set, and the following iteration convergence truncation criterion is configured:

[0130] ;

[0131] in and They represent the first time. Next and first In the next Newton iteration, there is a joint state vector containing unknown generalized coordinates and generalized momentum. This indicates that the Euclidean norm of the joint state vector is taken. This represents the preset residual convergence threshold, which takes the value of... It is on the constant order of magnitude.

[0132] At the same time, the system sets the maximum allowed number of iteration steps. If the iteration count reaches the maximum allowed number of iterations and the iteration convergence truncation criterion is still not met, the system performs a forced truncation operation and directly outputs the optimal solution of the previous control cycle as the input of the current control matrix to maintain the stable operation of the underlying physical frequency converter. After obtaining the optimal solution of generalized coordinates and the optimal solution of generalized momentum in the current convergent solution set, the system uses the Jacobian decoding matrix to inversely map the symplectic parameters in the high-dimensional space into the physical control quantities of the underlying device, and establishes the following decoding model:

[0133] ;

[0134] ;

[0135] in This represents the optimal discrete generalized coordinate solution extracted from the convergent solution set. This represents the discrete generalized momentum optimal solution extracted from the convergent solution set. This represents the Jacobian mapping matrix from the system's pre-configured generalized coordinates to the stator voltage. This represents the Jacobian mapping matrix from the generalized momentum to the stator voltage pre-configured in the system. This represents the stator voltage vector sequence extracted by decoding. This represents the variable frequency rate mapping gain matrix set by the system. This represents the frequency conversion rate sequence extracted by decoding.

[0136] The system combines the decoded stator voltage vector sequence with the frequency conversion rate sequence in matrix space to generate the final optimal control matrix that is directly issued.

[0137] By configuring an implicit Synlogen-Kutta integrator to perform high-order discretization and iterative solution of the continuous-time domain Hamiltonian canonical equations, the traditional explicit Euler integral algorithm overcomes the technical defect of introducing artificial numerical dissipation into the system due to the accumulation of truncation errors when calculating ultra-transient nonlinear extrema. This avoids the specific problem of electromechanical-hydraulic cross-interference tensor resonance pole excitation caused by model solution distortion during rapid frequency regulation of water pump station scheduling commands. The optimization process is strictly executed in the symplectic geometric operator space without artificial dissipation, ensuring that the output optimal control matrix approximates the true minimum dissipation boundary of the underlying physical equipment, forming a complete algorithm control closed loop.

[0138] The space vector inverter physical execution module is used to analyze the optimal control matrix to drive the variable frequency inverter devices of the water pumping station and execute physical variable frequency control actions.

[0139] Furthermore, the space vector inverted physics execution module is used to execute:

[0140] The optimal control matrix is ​​analyzed and decoded into a space vector pulse width modulation signal sequence.

[0141] Inject the space vector pulse width modulation signal sequence into the underlying drive channel of the frequency converter inverter device;

[0142] The on-duty cycle and off-duty cycle of the power switching elements in the frequency converter are controlled based on the space vector pulse width modulation signal sequence, and the space vector pulse width modulation signal sequence is converted into actual electromagnetic torque to perform physical frequency conversion control.

[0143] Specifically, the space vector inverter physical execution module undertakes the final execution task of dimensionality reduction mapping from the virtual computing space to the physical entity space in the digital twin system. The data input end of the space vector inverter physical execution module receives the optimal control matrix output by the preceding solution module. After the underlying pulse width modulation decoding and duty cycle reconstruction, the data output end generates the actual electromagnetic torque to drive the water pump unit, completing the unidirectional control link from digital command to mechanical work.

[0144] Analyzing the optimal control matrix requires extracting the target reference voltage vector and calculating the interaction time between adjacent non-zero base voltage vectors and the zero vector based on the volt-second balance equivalence principle. The system establishes the following analytical model for space vector pulse width modulation:

[0145] ;

[0146] ;

[0147] ;

[0148] in This indicates the duration of action of the first non-zero base voltage vector. This indicates the duration of action of the second non-zero base voltage vector. This indicates the duration of action of the zero vector. This represents the pulse width modulation switching period, which is usually set to a constant between 100 and 200 microseconds, depending on the main frequency of the underlying hardware. This represents the magnitude of the target reference voltage vector extracted from the analytical optimal control matrix. This indicates the measured value of the DC bus voltage inside the frequency converter inverter device. This represents the spatial electrical angle between the target reference voltage vector and the first non-zero base voltage vector.

[0149] After acquiring the operating time, the space vector inverter physical execution module converts it into a physical drive level and injects it into the underlying drive channel of the inverter device. Based on the generated space vector pulse width modulation signal sequence, the underlying drive channel strictly controls the on-time and off-time of internal power switching elements such as insulated-gate bipolar transistors. Under the influence of the duty cycle, the stator windings synthesize a rotating magnetic field, converting the electrical signal into physical mechanical work. This conversion process follows the following electromagnetic torque model:

[0150] ;

[0151] in This represents the actual electromagnetic torque generated by the final conversion. This indicates the number of pole pairs of the motor in the water pumping station. This represents the transient flux linkage component along the direct axis of the stator. This represents the transient flux linkage component of the stator quadrature axis. This represents the transient current component of the stator direct axis. This represents the stator quadrature-axis transient current component.

[0152] By directly decoding the optimal control matrix calculated purely mathematically into a space vector pulse width modulation signal sequence controlling the underlying actions of power switching elements, this method overcomes the traditional technical shortcomings of water pump station scheduling commands, which rely on macroscopic analog voltage open-loop issuance, leading to large electromagnetic torque pulsations and response lags. This avoids specific problems such as current exceeding limits or mechanical gear impact caused by hardware action delays when frequency converters execute high-speed frequency conversion commands. The virtual optimization matrix and physical hardware actions achieve strict microsecond-level synchronization within this module, ensuring the lossless reproduction of theoretical optimization results in a physical fluid environment.

[0153] The high-order harmonic residual dynamic closed-loop module is used to obtain the high-order harmonic residual vector generated when performing physical frequency conversion control actions, and feed the high-order harmonic residual vector back to the symplectic geometric electromechanical-hydraulic cross-coupled twin module to correct the system cross-coupled Hamiltonian energy function.

[0154] Furthermore, the high-order harmonic residual dynamic closed-loop module is used to perform:

[0155] During the execution of physical frequency conversion control, the physical stator high-order harmonic current components actually output to the motor by the frequency conversion inverter device are collected synchronously.

[0156] Extract the ideal reference high-order harmonic current component generated by the evolution prediction of the symplectic geometric electromechanical-hydraulic cross-coupled twin module under the same discrete time step;

[0157] High-frequency differential comparison calculations are performed on the high-order harmonic current components of the physical stator and the high-order harmonic current components of the ideal reference to generate high-order harmonic residual vectors.

[0158] The high-order harmonic residual dynamic closed-loop module is also used to perform:

[0159] The high-order harmonic residual vector is transmitted in reverse to the symplectic geometric electromechanical-hydraulic cross-coupled twin module.

[0160] A weighted compensation correction operator is constructed based on the higher harmonic residual vector;

[0161] The weight compensation correction operator is applied to the electromechanical-hydraulic cross-interference tensor to update the coupling weight coefficients inside the electromechanical-hydraulic cross-interference tensor.

[0162] The system cross-coupled Hamiltonian energy function for the next control cycle is reconstructed using the electromechanical-hydraulic cross-interference tensor after updating the coupling weight coefficients.

[0163] Specifically, the high-harmonic residual dynamic closed-loop module, acting as a reverse feedback data channel connecting the physical execution entity and the virtual twin computation space, is responsible for eliminating system time-varying errors and adaptively reconstructing the underlying physical boundaries. The module's data input receives high-frequency electromagnetic characteristics actually acquired by the underlying hardware and theoretically predicted characteristics from the twin space; its data output generates a corrected interference tensor and directly applies it to the energy operator in the next control cycle.

[0164] During the execution of physical frequency conversion control actions, the high-order harmonic residual dynamic closed-loop module extracts the physical stator high-order harmonic current components actually output by the frequency converter to the motor in real time. Simultaneously, the system uses the constructed system cross-coupled Hamiltonian energy function to solve for the partial derivatives of the flux linkage vector, calculates the time-domain reference current under the ideal state of completely undamped attenuation, and extracts high-frequency features using a fast Fourier transform. The derivation model is as follows:

[0165] ;

[0166] ;

[0167] in This represents the cross-coupled Hamiltonian energy function of the system before modification. This represents the generalized flux linkage state vector of the system. This represents the time-domain reference current sequence derived under ideal conditions. This represents the Fast Fourier Transform operator. This represents the high-pass filter mask matrix for extracting preset high-order harmonic frequency bands. This represents the extracted ideal reference higher harmonic current component.

[0168] After obtaining the above two sets of physical and theoretical comparison data, the system uses high-frequency differential comparison to calculate the degree of higher-order energy distortion caused by the aging and mechanical attenuation of the underlying physical magnetic circuit, and establishes the following residual extraction mathematical model:

[0169] ;

[0170] in This represents the calculated high-order harmonic residual vector. This represents the physical stator higher harmonic current component that the frequency converter actually outputs to the motor. This represents the ideal reference high-order harmonic current component generated by the evolution prediction of the symplectic geometric electromechanical-hydraulic cross-coupling twin module.

[0171] After acquiring high-frequency error data, the system transmits the higher harmonic residual vector in reverse to the twin modeling side. Based on the spatial distribution characteristics of the error matrix, a weight compensation correction operator is constructed to compensate and update the coupling weight coefficients inside the original electromechanical-hydraulic cross-interference tensor, establishing the following weight reconstruction model:

[0172] ;

[0173] ;

[0174] in This represents the weight compensation correction operator constructed based on the higher harmonic residual vector. This represents the preset error convergence gain matrix, which is a positive definite diagonal matrix containing fine-tuning constants. Indicates the current number Electromechanical-hydraulic cross-interference tensor used within the control cycle. This represents the electromechanical-hydraulic cross-interference tensor used in the next control cycle after updating the coupling weight coefficients.

[0175] The system utilizes the updated electromechanical-hydraulic cross-interference tensor to directly replace the original matrix and complete the reconstruction of the system cross-coupling Hamiltonian energy function for the next control cycle.

[0176] By extracting the high-order harmonic current components of the physical stator and constructing a high-order harmonic residual vector with ideal reference characteristics, and then updating the electromechanical-hydraulic cross-interference tensor in reverse, this method improves upon the traditional technical defects of digital twin control systems, which suffer from severe parameter drift between the virtual model and the physical entity due to mechanical wear and magnetic circuit aging caused by long-term equipment operation. This avoids the specific problems of divergence in underlying optimization commands and secondary increases in system energy consumption caused by twin mapping distortion in water pumping stations under long-cycle control conditions. The reverse closed-loop feedback mechanism ensures that the solution boundary in the pure mathematical space strictly conforms to the actual physical decay state of the physical hardware, guaranteeing the closed-loop reliability of the ultra-transient optimization control algorithm throughout its entire lifecycle.

[0177] Example 2:

[0178] In applications such as large-scale municipal cascade water supply networks and long-distance inter-basin water diversion projects, water pumping stations often need to cope with sudden pipe bursts or extreme water demand peaks that cause drastic changes in the water demand topology of the wide-area network. Faced with such ultra-transient conditions, the dispatching system must command multiple pumping units within the station to perform ultra-fast frequency conversion control. A serious technical problem exists in this application scenario: traditional water dispatching architecture calculates the electromagnetic transient evolution model of the driving source separately from the fluid dynamics transient damping model of the load side. When the voltage vector output by the frequency converter undergoes a rapid change, it easily leads to saturation of the motor stator and rotor flux linkages and high-order harmonic electromagnetic losses; this transient electromagnetic torque further excites nonlinear water hammer pressure waves within the manifold space of the pipeline network. The overlap of the physical frequencies of the electromagnetic transient high-frequency oscillations and the fluid transient pressure waves induces strong transient electromechanical-hydraulic resonances. This problem directly leads to the risk of motor overcurrent tripping and accelerated cavitation and spalling of the pump impeller during ultra-transient frequency conversion, damaging the operational stability of the underlying hardware and generating extreme transient dynamic energy dissipation. To solve these problems, this invention provides an intelligent energy-saving scheduling system for water pumping stations based on digital twins, the structure of which is as follows: Figure 1 As shown. The specific implementation process of this system is as follows:

[0179] A cross-physics-field underlying control closed-loop architecture was established to address the ultra-transient variable frequency operation of water pumping stations. The ultra-transient heterogeneous tensor high-frequency sensing module synchronously acquires alternating motor and fluid data from the hardware source and outputs the corresponding electromagnetic and fluid transient state vectors in a reduced-dimensional manner. The symplectic geometric electromechanical-hydraulic cross-coupling twin module calculates the electromechanical-hydraulic cross-interference tensor based on these vectors and introduces anti-cavitation differential constraint rules to construct the system's cross-coupling Hamiltonian energy function, completing the energy mapping of physical entities to a unified symplectic manifold computational space. The nonholonomic constraint symplectic optimization control module uses a symplectic integral algorithm to iteratively solve this energy function without numerical dissipation, outputting an optimal control matrix that minimizes energy dissipation. The space vector inverter physical execution module directly parses this matrix to drive the underlying variable frequency inverter devices to perform physical frequency conversion actions. The high-order harmonic residual dynamic closed-loop module extracts the high-order harmonic residual vector generated on the execution side and uses iterative feedback to correct the Hamiltonian energy function. This technical solution breaks through the limitations of the traditional digital twin architecture, which separates electromagnetic transient and hydraulic transient models. It achieves extreme value optimization across physical fields within the Bauschian geometric framework, avoiding motor overcurrent tripping and impeller cavitation and spalling problems caused by electromechanical-hydraulic cross resonance during high-speed frequency regulation of pump station units. It ensures the closed-loop anti-disturbance capability and low-dissipation execution of control commands throughout the entire life cycle.

[0180] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart energy-saving scheduling system for water pumping stations based on digital twins, characterized in that, Includes the following modules: The ultra-transient heterogeneous tensor high-frequency sensing module is used to collect motor data and fluid data of the water pumping station and output the electromagnetic transient state vector of the motor and the transient state vector of the fluid. The symplectic geometric electromechanical-hydraulic cross-coupling twin module is used to receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid, calculate the electro-hydraulic cross-interference tensor, and construct the system cross-coupling Hamiltonian energy function in combination with the preset anti-cavitation differential constraint rules. The nonholonomic constraint symplectic optimization control module is used to iteratively solve the cross-coupled Hamiltonian energy function of the system using a symplectic integral algorithm and output the optimal control matrix. The space vector inverter physical execution module is used to analyze the optimal control matrix to drive the variable frequency inverter device of the water pumping station and perform physical variable frequency control actions. The high-order harmonic residual dynamic closed-loop module is used to obtain the high-order harmonic residual vector generated when the physical frequency conversion control action is executed, and feed the high-order harmonic residual vector back to the symplectic geometric electromechanical-hydraulic cross-coupled twin module to correct the cross-coupled Hamiltonian energy function of the system.

2. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 1, characterized in that, The ultra-transient heterogeneous tensor high-frequency sensing module is used to perform: Collect the stator transient current data and rotor transient flux data of the frequency converter device of the water pumping station in the three-phase static coordinate system; A coordinate transformation operation is performed on the stator transient current data and the rotor transient flux data to map the stationary coordinate system data to the synchronous rotating coordinate system; The direct-axis transient current component, the quadrature-axis transient current component, the direct-axis transient flux component, and the quadrature-axis transient flux component are extracted respectively in the synchronous rotating coordinate system. The direct-axis transient current component, the quadrature-axis transient current component, the direct-axis transient flux linkage component, and the quadrature-axis transient flux linkage component are tensor-concatenated to generate the electromagnetic transient state vector of the motor.

3. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 1, characterized in that, The ultra-transient heterogeneous tensor high-frequency sensing module is also used to perform: Simultaneously collect transient outlet flow data of multiple water pumps at the water pumping station and transient node water pressure fluctuation data of key topology nodes in the wide area pipeline network; Extract the first-order time derivative of the transient outlet flow rate data of the water pump and the transient node water pressure fluctuation data within the rolling time window; The transient outlet flow rate data of the water pump, the transient node water pressure fluctuation data, and the first-order time derivative are reconstructed into tensors according to a preset dimension to generate the transient state vector of the fluid.

4. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 1, characterized in that, The symmetric geometric electromechanical-hydraulic cross-coupled twin module is used to perform: Receive the electromagnetic transient state vector of the motor and the transient state vector of the fluid; Establish a symplectic manifold space for mapping the energy evolution characteristics of the system, and extract a Riemannian metric tensor that matches the transient state vector of the fluid and characterizes the transient hydraulic damping characteristics of the pipe network in the symplectic manifold space; Extract the initial electromechanical interference matrix characterizing the physical coupling relationship between the electromagnetic torque of the motor and the fluid reaction torque of the pump impeller; The Riemannian metric tensor is injected as a nonlinear damping feature into the initial electromechanical interference matrix, and the electromechanical-hydraulic cross-interference tensor is generated by mapping.

5. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 4, characterized in that, The symmetric geometric electromechanical-hydraulic cross-coupled twin module is also used to perform: Extract the anti-cavitation differential constraint rules that reflect the transient rate of fluid state vector triggering local cavitation boundaries; Based on the law of conservation of energy, an electromagnetic inertia penalty term is established using the electromagnetic transient state vector of the motor, and a hydraulic damping penalty term is established using the fluid transient state vector and the Riemann metric tensor. The electromagnetic inertial penalty term, the hydraulic damping penalty term, the electromechanical-hydraulic cross-interference tensor, and the anti-cavitation differential constraint rule are functionally aggregated in the symplectic manifold space to generate the system cross-coupled Hamiltonian energy function.

6. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 4, characterized in that, The nonholonomic constraint symplectic optimization control module is used to execute: Receive the cross-coupled Hamiltonian energy function of the system; Establish a symplectic space for system integration, and integrate and map the electromagnetic transient state vector of the motor and the transient state vector of the fluid into a generalized coordinate vector and a generalized momentum vector in the symplectic space for system integration. Using the generalized coordinate vector and the generalized momentum vector, a set of continuous-time-domain Hamiltonian canonical equations corresponding to the cross-coupled Hamiltonian energy function of the system is established.

7. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 6, characterized in that, The nonholonomic constraint symplectic optimization control module is also used to execute: Configure the Bausch integral algorithm as an implicit Singlonga-Kutta integrator; The implicit Singlonga-Kutta integrator is used to perform high-order discretization of the continuous-time domain Hamiltonian canonical equations to generate an internal nonlinear node constraint equation system. Under the constraint boundary of maintaining the conservation of the symplectic geometric area element, the Newton iteration method is used to iteratively solve the constraint equations of the internal nonlinear nodes to extract the stator voltage vector sequence and the frequency conversion rate sequence. The optimal control matrix is ​​generated by combining the stator voltage vector sequence with the frequency conversion rate sequence.

8. The intelligent energy-saving scheduling system for water pumping stations based on digital twins according to claim 1, characterized in that, The space vector inverted physics execution module is used to execute: The optimal control matrix is ​​analyzed and decoded into a space vector pulse width modulation signal sequence; The space vector pulse width modulation signal sequence is injected into the underlying drive channel of the frequency inverter device; The on-time and off-time of the power switching elements in the frequency converter are controlled according to the space vector pulse width modulation signal sequence, and the space vector pulse width modulation signal sequence is converted into actual electromagnetic torque to perform the physical frequency conversion control action.

9. A smart energy-saving scheduling system for water pumping stations based on digital twins according to claim 1, characterized in that, The high-order harmonic residual dynamic closed-loop module is used to perform: During the execution of the physical frequency conversion control action, the physical stator high-order harmonic current components actually output by the frequency conversion inverter device to the motor are simultaneously collected; Extract the ideal reference high-order harmonic current component generated by the evolution prediction of the symplectic geometric electromechanical-hydraulic cross-coupled twin module under the same discrete time step; The high-order harmonic current components of the physical stator and the high-order harmonic current components of the ideal reference are compared using high-frequency differential calculation to generate the high-order harmonic residual vector.

10. A smart energy-saving scheduling system for water pumping stations based on digital twins according to claim 9, characterized in that, The high-order harmonic residual dynamic closed-loop module is also used to perform: The higher harmonic residual vector is transmitted in reverse to the symplectic geometric electromechanical-hydraulic cross-coupled twin module; A weighted compensation correction operator is constructed based on the aforementioned higher harmonic residual vector; The weight compensation correction operator is applied to the electromechanical-hydraulic cross-interference tensor to update the coupling weight coefficients inside the electromechanical-hydraulic cross-interference tensor; The system cross-coupled Hamiltonian energy function for the next control cycle is reconstructed using the electromechanical-hydraulic cross-interference tensor after updating the coupling weight coefficients.

Images