236 results about "Radial basis function neural" patented technology

A robot kinetic parameter identification method based on double-layer iteration and friction compensation comprises the following steps: S1, establishing a kinetic model of a robot, and performing linearization processing on the kinetic model to obtain a linearization model represented by an observation matrix and an inertial parameter vector; s2, designing an excitation trajectory for the linearized model by adopting improved Fourier series of a quintic polynomial, and setting constraint conditions of joint positions, speeds and accelerated speeds; s3, the robot is controlled to move according to the excitation track, and joint state data of the robot are collected and subjected to noise reduction processing; and defining the noise-reduced driving torque as a measurement torque. According to the robot kinetic parameter identification method based on double-layer iteration and friction compensation, the physical feasibility of the robot kinetic parameters can be ensured, the friction model is improved to identify the friction parameters, the friction model is fitted by adopting the radial basis function neural network, and the precision of subsequent robot control is ensured.

A system, computer readable storage medium and method for controlling a trajectory of coordinated time-varying maneuvers of a geometric formation of unmanned vehicles is disclosed. The system includes unmanned vehicles, each configured with communication circuitry to communicate between the vehicles. A subset of the unmanned vehicles function as leader vehicles, with the remaining vehicles functioning as follower vehicles for leader-follower maneuvering. The system further includes an actuator suite configured to adjust the direction and orientation of each vehicle, a sensor suite for stabilization and navigation, and a flight controller for maintaining stable maneuvering, even in the presence of actuator faults and sensor deception attacks. Processing circuitry is configured with a reinforcement learning neural network that includes identifier, actor, and critic radial basis function neural networks to estimate movement, adjust control actions, and assess vehicle performance based on feedback signals, including corrupted signals from the sensor suite due to deception attacks.

The invention provides an unmanned aerial vehicle-unmanned vehicle combined formation cooperative control method and system, and relates to the technical field of vehicle-vehicle cooperation, and the method comprises the steps: taking a virtual unmanned aerial vehicle as a leader, taking the unmanned aerial vehicle and the unmanned vehicle as followers in a unified manner, setting a fixed formation offset for each follower, and defining a communication topological relation based on a graph theory; the method comprises the following steps: acquiring an actual measurement state vector of a sensor in real time, processing a current estimation state through a constructed radial basis function neural network observer, outputting a disturbance estimation value in combination with a weight matrix, updating the estimation state according to a core equation, and judging whether to update the weight matrix or not according to an observation error; calculating the expected state of the follower according to the reference trajectory of the leader and the fixed formation offset, calculating the formation error, constructing an event trigger communication condition, judging whether the condition is met or not, enabling the follower to interact the state and the error according to the communication topology only when the condition is met, or else, continuing to use the previous trigger data; and finally, current control input of the follower is calculated through a distributed control law.

The invention relates to the technical field of robot control, in particular to a cross-modal perceptual driving body robot compliance control system which comprises the steps that a sensor is adopted to synchronously collect environment information, multi-source data bias is eliminated through a space-time alignment algorithm, a radial basis function neural network is adopted to analyze multi-modal fusion features, and a multi-modal model is obtained; human operation intention probability distribution is extracted to decompose a task into a path planning layer and a motion control layer, a collision-free trajectory is generated through an RRT algorithm, a high-fidelity physical engine is adopted to construct a virtual interaction scene, and robot learning results are shared through federal learning. According to the method, the problems of inaccurate perception, incoordination between intention recognition and interaction control, difficulty in control strategy verification, slow new task adaptation and difficulty in multi-robot learning result sharing caused by multi-source data deviation and large multi-modal semantic difference of the body robot in a complex environment are solved.

The invention discloses a self-adaptive brushless motor control method and system, and relates to the field of intelligent control, and the method comprises the steps: collecting the original data of the operation state of a motor through a sensor group, and carrying out the preprocessing; time-varying parameter identification is completed through combination of an extended Kalman filtering algorithm and a radial basis function neural network, an evaluation index system is constructed based on an analytic hierarchy process to obtain a comprehensive evaluation value, and a related trend is predicted through a long and short-term memory neural network; constructing a multi-modal control strategy library, determining an adaptive strategy, optimizing core parameters by using an improved particle swarm optimization algorithm, generating a control instruction, and outputting a corresponding current through a power driving module; and monitoring motor parameters in real time, comparing with a control target value, calculating deviation, correcting an identification result, adjusting a strategy threshold value, and updating and optimizing an objective function. The method has the advantages that by accurately sensing the state of the motor, dynamically adapting the control strategy and optimizing parameters in real time, it is ensured that the motor stably and efficiently operates under the complex working condition, and the characteristics of energy conservation and long service life are achieved.

The invention discloses a J-A model parameter identification method, system, equipment and medium based on RBF and an improved brownish bear algorithm, and belongs to the technical field of power system optimization, and the method comprises the steps: building a Jiles-Atherton hysteresis reverse model of a current transformer, determining a to-be-identified parameter vector, and building a model with a root-mean-square error between actually measured magnetic field intensity and simulated magnetic field intensity as a target function, training a radial basis function neural network model, expanding data through linear interpolation processing, obtaining a predicted magnetic induction intensity value, inputting an objective function and radial basis function prediction data into an improved brownish bear optimization algorithm, and iteratively optimizing model parameters through hierarchical population position updating and fitness evaluation until convergence conditions are met. And outputting an optimal parameter identification result. According to the method, high-precision and high-efficiency identification of hysteresis model parameters is realized, the generalization capability and robustness of the system are improved, and reliable technical support is provided for hysteresis characteristic analysis of a complex physical system.

The invention discloses a modular mechanical arm actuator additive fault optimal fault-tolerant control method and device of a zero sum differential game, and relates to the field of robot control algorithms, and the method comprises the steps: representing a nonlinear damping characteristic through a joint friction torque, describing the dynamic interaction of multiple joints through a cross-linking coupling item, and determining the optimal fault-tolerant control of the additive fault of the modular mechanical arm actuator; a dynamical model containing faults is constructed. And uncertain items in the model are updated online by adopting a radial basis function neural network identifier, so that the model precision is improved. A performance index function is constructed based on position errors, actuator faults and controller input are regarded as two opposite parties of a zero and differential game, the performance index function is approximated through a single evaluation neural network, a Hamiltonian-Jacobi-Axaxi equation is approximately solved, and an optimal fault-tolerant control strategy is obtained. According to the method, the game theory is combined with the neural network, the dynamic unknown fault problem of the modular mechanical arm is effectively solved while the system energy consumption integration is reduced, and real-time optimal control over the modular mechanical arm is achieved.

The invention relates to a method and device for predicting online open course learner satisfaction and electronic equipment, and the method comprises the steps: predicting the online open course satisfaction of students through an MLP and RBF neural network model by using a virtual learning environment of a large-scale online teaching and learning platform and combining learning behavior data in a learning management system (LMS); the model comprises a data acquisition and processing module, a multilayer perceptron (MLP) module, a radial basis function (RBF) neural network module, a classification tree module and a control block, the data acquisition and processing module is used for generating a training and testing data set, and the MLP and RBF neural network model predicts the satisfaction degree of a learner. The MLP model carries out feature extraction through a multi-layer perceptron structure and different activation functions, the RBF model measures the distance between input data and a center by using a radial basis function to realize feature extraction, the classification tree is used for judging a prediction model to which a data point belongs, the control block integrates features from the MLP and RBF neural network models, and the RBF model is used for determining a prediction model to which the data point belongs. Experimental results show that the prediction accuracy of low-satisfaction-degree learners and high-satisfaction-degree learners can be improved at the same time through the combination scheme of the MLP and the RBF, the method can be applied to learner satisfaction degree prediction of various online open courses, an educational institution is helped to know the satisfaction degree condition of students in time, course design and teaching strategies are optimized, and the teaching efficiency is improved. And important support is provided for teaching reform and optimization in the field of online education.

The invention discloses a mechanical arm trajectory tracking algorithm based on an improved sliding mode adaptive neural network, and belongs to the technical field of mechanical arm control. The method comprises the steps that recorded data of an improved sliding mode and a neural network about the mechanical arm and estimated values under the influence of actual modeling errors and external disturbance are obtained, and a mechanical arm dynamic model is jointly constructed; through the collaborative design of dynamic modeling optimization, sliding mode control improvement and adaptive RBF neural network compensation, the three core problems of low precision, large buffeting and poor reliability in mechanical arm trajectory tracking are synchronously solved; the radial basis function neural network is adopted for online estimation, the influence of modeling errors and external unknown disturbance on a mechanical arm system is avoided, the trajectory tracking precision is improved, and meanwhile it is guaranteed that all joint angles operate stably, and sudden change of torque does not exist.

The invention discloses an unmanned ship adaptive optimal interference control method based on reinforcement learning, and particularly relates to the technical field of unmanned ship automatic control, and the method comprises the steps: building an unmanned ship trajectory kinematics model based on the position information and heading angle information of an unmanned ship under a geodetic coordinate system, and the corresponding speed information under an unmanned ship appendage coordinate system; the method comprises the following steps: establishing an unmanned ship trajectory dynamics model by considering the control operation of the unmanned ship and the time-varying environment interference problems of wind, waves, flow, unmodeled dynamics and the like in a marine environment in which the unmanned ship is located; introducing a radial basis function neural network based on a set unmanned ship trajectory mathematical model; designing a self-adaptive interference observer to estimate and offset time-varying environment interference in unmanned ship trajectory tracking; based on a radial basis function neural network and an interference observer, an unmanned ship adaptive interference suppression controller is designed by using an adaptive vector backstepping method.

The invention belongs to the technical field of power distribution automation of a power system, and particularly relates to a self-adaptive load control method and system for an intelligent cable branch box, and the method comprises the steps: collecting and preprocessing the current data of a branch loop and the temperature data of a cable joint in real time, obtaining a data confidence factor, and extracting a current change rate and a temperature rise rate; obtaining a load prediction value in a future set time period through a load trend prediction model; calculating the residual heat tolerance time for reaching the limit tolerance temperature; calculating an output load control adjustment coefficient based on the adaptive load control decision model and a particle swarm optimization algorithm; and judging a current state interval according to the load control adjustment coefficient, and controlling an execution end to execute a grading response strategy of through-flow maintenance, short-time overload monitoring and early warning and breaking current limiting in combination with the residual heat tolerance time. According to the method, the residual heat tolerance time is calculated in combination with load prediction and a transient thermal circuit model, and hierarchical response of load control is realized through a particle swarm optimized radial basis function neural network decision.

The invention provides a flying vehicle path planning method based on dynamic weight adjustment, and belongs to the technical field of aircraft planning. Comprising the following steps: acquiring planning data under different fixed weight values; processing planning data according to actual vehicle energy storage and task limiting time, and fitting weight mathematical representations under different task requirements by using a radial basis function neural network RBFNN (Radial Basis Function Neural Network); the method comprises the following steps: designing a weight A * algorithm of dynamic weight adjustment, dynamically adjusting weight values of different costs, searching a path node with the minimum comprehensive cost, reasonably switching different motion modes and planning a short-time energy-saving task path on the basis of weight mathematical representation and aiming at task requirements changing in real time. According to the method, reasonable switching of different motion modes is realized by searching the path node with the minimum comprehensive cost, and a short-time energy-saving task path is planned. Weight mathematical representations under different task requirements are constructed through the RBFNN, and weight values of different costs are dynamically adjusted to meet the task requirements changing in real time.

The invention discloses an advanced dynamic evaluation method for dolomite sanding characteristics. The advanced dynamic evaluation method comprises the following steps: determining mesoscopic and macroscopic thresholds of dolomite sanding degree grading as rock test static parameters based on an indoor test; extracting kinetic parameters and electrical parameters of the dolomite rock mass as advanced detection dynamic parameters based on geophysical advanced detection of missile-electricity combination; performing principal component analysis and correlation analysis on the rock test static parameters and advanced detection dynamic parameters, and constructing a dolomite sanding dynamic-static parameter quantitative conversion model by using RBFNN (Radial Basis Function Neural Network); training an intelligent grading evaluation model by taking the rock test static parameters as input and the sanding degree grade as output; and mapping the advanced detection dynamic parameters into rock test static parameters through the quantitative conversion model, inputting the mapped rock test static parameters into the intelligent grading evaluation model for verification and iterative optimization, and outputting a sanding degree grading result. Advanced prevention and control of sanding of the dolomite in front of the tunnel construction face can be achieved, and the ground disaster of sand collapse caused by sanding and sand gushing is avoided.

The invention discloses a time sequence reconstruction and machine learning coupled chlorophyll monitoring data breakpoint restoration method, and belongs to the technical field of water quality monitoring. The invention discloses a chlorophyll monitoring data breakpoint restoration method based on coupling of time sequence reconstruction and machine learning, and the method comprises the following steps: S1, collecting water quality monitoring data, and cleaning the monitoring data to obtain preprocessed data; s2, performing time sequence reconstruction on the preprocessed data to obtain a weekly average 1 data set; s3, respectively constructing a radial basis function neural network model and a back propagation neural network model by taking the chlorophyll concentration as a response variable and the conventional water quality parameter as a predictive variable; s4, performing performance evaluation on each model by taking a root mean square error, an average absolute percentage error, goodness of fit and relative error distribution statistics as evaluation indexes, and screening out an optimal model; and S5, applying the conventional water quality parameters in the breakpoint interval of the chlorophyll monitoring data in the water body to the optimal model, and outputting the restored chlorophyll concentration value to complete the dynamic restoration of the breakpoint.

The invention belongs to the technical field of CFD (computational fluid dynamics) parameter calibration, and discloses a CFD parameter adaptive calibration method and system based on measured data and a double-agent model, and the method comprises the steps: obtaining a CFD input parameter sample, inputting the CFD input parameter sample into a CFD solver, and obtaining an initial simulation result; determining an error evaluation index according to the initial simulation result based on a target actual measurement data result; constructing a double-agent model based on a Kriging model and a radial basis function neural network by taking a CFD input parameter sample as an independent variable and an error evaluation index as a dependent variable; the double-agent model is trained, the trained double-agent model takes the error evaluation index as fitness, and CFD input parameter values are obtained based on a genetic algorithm; the CFD input parameter values are input into the CFD solver for a simulation experiment, a calibrated simulation result is output, the reliability and generalization ability of prediction are improved through a double-agent model, a high-fidelity simulation result is output through the CFD solver, and the number of times of calling the CFD solver is reduced while the calibration precision is guaranteed.

The invention discloses a geophysical method for predicting coal rock thickness, particularly relates to the technical field of coalbed methane exploration, and has the core innovation that a wedge-shaped geologic model containing nine kinds of roof and floor lithology combinations is constructed, and seismic attributes sensitive to a thin coal seam are optimized through principal component analysis; converting the complex relationship between the seismic attributes and the coal thickness into a linear separable problem by using the specific high-dimensional nonlinear mapping capability of the radial basis function neural network; a three-level closed-loop mechanism of attribute optimization-network training-dynamic verification is established, and model self-optimization is realized through spider diagram analysis, over-fitting monitoring and new well triggering iteration. According to the method, the tuning distortion problem of a traditional linear model in thin coal seam prediction is solved, the reliability of well-free area prediction is remarkably improved, and a key technical support is provided for coal seam gas dessert identification and development decision making.

The invention belongs to the technical field of hydraulic engineering digital twinning, and provides an arch dam static response rapid calculation method which comprises the following steps: sampling and simulating based on a finite element model in an offline stage to obtain a displacement field snapshot matrix, extracting a dominant mode through intrinsic orthogonal decomposition to construct a reduced-order subspace, and calculating a mode coefficient; and training a radial basis function neural network to establish a nonlinear mapping relation by taking the working condition parameters as input and the modal coefficient as output, and finally, inputting the target working condition parameters into the trained network to predict the modal coefficient in an online stage, and quickly reconstructing a complete displacement field response through linear combination with the POD modal. The method achieves the quick and accurate prediction of the static physical field of the arch dam, and remarkably improves the calculation efficiency of the displacement field of the arch dam.

The invention discloses a piezoelectric micro-positioning platform preset performance control method based on an extended state observer, and the method mainly comprises the steps: firstly, building a piezoelectric micro-positioning platform system model considering input hysteresis and unknown disturbance; secondly, designing an extended state observer based on a radial basis function neural network to estimate an unmeasurable state of the system and lumped disturbance containing input hysteresis; then, a preset performance function is used, so that the overshoot performance of the system is effectively improved; thirdly, in combination with a preset performance function, a first-order sliding mode differentiator and an expansion state observer, providing a virtual control law and an adaptive control law; and finally, a preset performance controller is utilized, a Lyapunov stability theory is combined, proper parameters are selected to ensure that the closed-loop system is kept stable within preset time, and tracking errors are controlled within a set error range.

The invention provides an intelligent fireproof and high-temperature steam cleaning oil fume integrated hood, and relates to the technical field of range hoods, and the intelligent fireproof and high-temperature steam cleaning oil fume integrated hood comprises an oil fume diffusion monitoring module used for obtaining an oil fume image; through a difference image method, according to the difference between the oil smoke image and the reference image, obtaining oil smoke diffusion data; vOC data are obtained; calculating a lampblack diffusion quantized value; the intelligent cleaning module is used for determining the optimal steam release temperature and the optimal steam release amount through an improved artificial neural network; according to the optimal steam release temperature and the optimal steam release amount, the steam release amount and temperature of the range hood are dynamically adjusted based on a PID control system of a radial basis function neural network, and the smoke hood is cleaned by releasing high-temperature steam; the flame monitoring module is used for acquiring a cooking bench image through a camera; through a convolutional neural network, identifying a flame combustion degree, and determining a kitchen fire occurrence probability value; and the fire extinguishing module is used for spraying water to extinguish the kitchen.

The invention relates to the technical field of artificial intelligence and control, in particular to a non-linear multi-agent system fixed time consistency control method and device and a medium. According to the control method, a backstepping recursion method is used as a control design framework, a self-adaptive fixed time consistency controller based on an auxiliary compensation system is provided, and the problem of non-linear multi-self-body system consistency control under time-varying input delay is solved. By providing and changing a control input signal by a person, the motion trajectory of the multi-agent system can be modified as required. Time-varying input time delay and an unknown nonlinear function are considered in a system model, so that the system is more general. Wherein an unknown time-varying delay function is processed by constructing an auxiliary compensation system, and an unknown nonlinear function is processed by using radial basis function neural network approximation. According to a fixed time control correlation lemma, an actual fixed time adaptive consistency control method is provided to ensure that the consistency error of the multi-agent system converges to a neighborhood near an original point in fixed time.

The present application relates to the technical field of aquaculture, and particularly relates to an improved PSO-based water quality parameter prediction method and system for aquaculture, which comprises collecting water quality parameters at different positions and depths in a breeding pond; training an improved radial basis function (RBF) neural network using training set data; and optimizing the parameters of the improved RBF neural network model using an improved particle swarm optimization (PSO) algorithm. The present application introduces a mixed Gaussian function and an abnormal S-shaped function into the radial basis function of the traditional RBF neural network, thereby solving the problem of weak capability of the model in nonlinear data modeling. Furthermore, the present application improves the inertia factor and the learning factor in the traditional PSO algorithm, thereby solving the problems of slow parameter convergence speed and poor global search capability in the RBF neural network.

The invention relates to the technical field of vehicle state monitoring, in particular to a commercial vehicle re-recognition method and device based on a neural network and a medium, and the method mainly comprises the steps: constructing a radial basis function neural network model which comprises an input layer, a hidden layer and an output layer; optimizing parameters of the radial basis function neural network model by using a Bayesian optimization algorithm to obtain an optimal diffusion constant and an optimal regularization coefficient, and then adjusting hyper-parameters of the radial basis function neural network model; and collecting real-time operation data of a vehicle, and inputting the real-time operation data into the pre-trained radial basis function neural network model according to a time sequence to obtain a load identification result. The method overcomes the defects that a model based on a dynamic formula is difficult in parameter calibration and insufficient in stability and generalization ability.

The invention discloses a multi-under-actuated AUV distributed predefined time formation tracking control method and system based on data driving and a medium, and belongs to the field of formation control of unmanned underwater vehicles. A control strategy under an adjustable predefined time stability framework is provided; convergence time can be set by user priori and is irrelevant to initial conditions, and actual stabilization time is adjusted through designable parameters; in order to eliminate dependence on prior model information, complete data driving modeling based on a radial basis function neural network is adopted, and unknown dynamics is reconstructed only by using system input and output data; in order to suppress high-frequency buffeting caused by rapid convergence, a neural shunt mechanism is introduced to realize sliding mode switching smoothing and dynamic suppression; and constructing a distributed three-dimensional formation tracking control law based on the consistency error. According to the method, convergence time can be controlled, buffeting and control input peak values are remarkably reduced, formation tracking precision and robustness are improved, actuator loss is reduced, and the method is suitable for a multi-AUV cooperative task in a complex marine environment.

The invention discloses a battery static maintenance time prediction method and system, and mainly relates to the technical field of vehicle storage battery state monitoring and prediction. Comprising the following steps: collecting battery parameters of a vehicle at multiple power-off moments and power-on moments, and collecting data in a classified manner according to an electric switch state of a storage battery power supply 30; based on the collected battery parameter data, training a self-feedback radial basis function neural network (RBFNN) model; during the power-off period of the vehicle, the battery parameters at the last power-off moment are read regularly, the corresponding trained RBFNN model is selected according to the state of the 30-electric switch, and the battery parameters are input to predict the static maintenance time; and calculating the time difference from the current time to the last power-off time and the remaining holding time, comparing the remaining holding time with a preset time threshold value, and sending charging reminding information to the user through the cloud system. The method has the beneficial effect that the problems of battery loss and incapability of starting the vehicle due to long-term static state of the vehicle can be effectively avoided.

The invention belongs to the technical field of magnetic suspension control, and provides a multi-point cross-coupling suspension control method for a sliding-mode-driven RBF network, and the method comprises the steps: constructing a sliding-mode surface and a radial basis function neural network according to the suspension gap error of a single-point electromagnet, and obtaining a sliding-mode-driven RBF neural network control method for the single-point electromagnet; independent sliding mode driving RBF neural network control is respectively carried out at four electromagnets, gap and speed dual cross coupling terms are designed according to errors among points in a four-point suspension frame, error terms are superposed in control input, and finally a sliding mode surface driving RBF controller combined with cross coupling control is formed. According to the method, the tracking performance of the target gap and the synchronization performance between the suspension points are improved, stable suspension of the suspension frame is achieved, and it is verified that the method still has a remarkable control effect under the operation conditions that the track is not smooth and faults occur.

The invention relates to the technical field of maglev train suspension control, in particular to a flexible suspension module control method and system for a maglev train, which can remarkably improve the tracking precision, response speed and anti-interference capability of a suspension gap, effectively suppress flexible vibration and ensure high performance and high reliability of a suspension system. The invention provides a flexible suspension module control method for a maglev train. The flexible suspension module control method comprises the following steps: S1, constructing a suspension module coupling dynamic model comprising a flexible electromagnet and a flexible F rail; s2, designing a super-spiral sliding mode-radial basis function neural network adaptive controller; s3, performing anti-saturation correction on the output of the controller; and S4, converting the corrected control quantity into expected current, and controlling a power switch to drive an electromagnet to execute suspension control through a driving circuit.

The invention discloses a robot anti-interference control method based on a neural network interference observer, and the method specifically comprises the following steps: 1, building a robot dynamics model containing external interference, and constructing a state space model; step 2, establishing a fixed time interference observer based on a radial basis function neural network, and performing external interference estimation on the robot kinetic model; and step 3, a fixed time sliding mode controller is designed based on the interference estimation result, and fixed time control under the condition of external interference is realized. According to the robot anti-interference control method based on the neural network interference observer, the interference is accurately and rapidly estimated by using the fixed time interference observer, the convergence of the system state in the fixed time is realized by combining the fixed time sliding mode controller, and the feedforward compensation and feedback suppression of the interference are combined, so that the system stability is improved. The method has the advantages of simple implementation mode, high interference estimation precision and high response speed.

A method for controlling a fuel cell hybrid electric vehicle (FCHEV) with sensor fault tolerance includes receiving current measurements of a fuel cell, a battery, an ultracapacitor, and a voltage measurement of a DC-bus from sensors; detecting and estimating a sensor fault in at least one of the current measurements using a radial basis function neural network (RBFNN); calculating a new value for a single parameter of the RBFNN according to a minimum learning parameter scheme; calculating a duty cycle value for each power converter; and applying the calculated duty cycle value to power converters that connect the fuel cell, the battery, and the ultracapacitor to the DC-bus, to maintain a current distribution despite the sensor fault.

According to the fault-tolerant control method for the actuator fault in the urban sewage treatment process, stable control over the dissolved oxygen concentration and the nitrate nitrogen concentration in the urban sewage treatment process under the influence of the actuator fault is achieved. Firstly, a nonlinear observer is designed to accurately estimate an actuator fault; secondly, constructing a radial basis function neural network to accurately characterize the time-varying nonlinear dynamic characteristics of the sewage treatment process; a fault-tolerant controller based on self-adaptive dynamic programming is designed, and the problem that the dissolved oxygen concentration and the nitrate nitrogen concentration deviate from set values due to actuator faults is solved jointly by constructing an evaluation network and an execution network, training a neural network by adopting a strategy iterative algorithm, solving an approximate optimal control law and fusing a prior classical controller. Experimental results show that the method can realize fault-tolerant control of the dissolved oxygen concentration and the nitrate nitrogen concentration, and ensures stable operation of the urban sewage treatment process.

Provided are a method and a device for reconstructing lost data of marine wireless sensor networks (MWSNs). The data reconstruction method includes following steps: establishing an initial topological structure of the MWSNs; using an improved hierarchical energy balance multipath (IHEBM) routing protocol to cluster network nodes; using an improved radial basis function neural network (RBFNN) to predict lost data of nodes in the cluster based on a clustering of the nodes; and using a centralized principal component analysis (PCA) method to compress and reconstruct data of a cluster head node in a process of data transmission from a cluster head to a ship base station.