84 results about "Neural fuzzy" patented technology

System for determining the occupancy of a seat in a vehicle using a variety of transducers and pattern recognition technologies and techniques that applies to any combination of transducers that provide information about seat occupancy. These include weight sensors, capacitive sensors, inductive sensors, ultrasonic, optical, electromagnetic, motion, infrared, and radar among others. The system includes a processor coupled to the transducers for receiving the data from the transducers and processing the data to obtain an output indicative of the current occupancy state of the seat. An algorithm is resident in the processor and is created from a plurality of data sets, each representing a different occupancy state of the seat and being formed from data from the transducers while the seat is in that occupancy state. The algorithm produces the output indicative of the current occupancy state of the seat upon inputting a data set representing the current occupancy state of the seat and being formed from data from the transducers. The algorithm may be a neural network or neural fuzzy algorithm generated by an appropriate algorithm-generating program.

The invention provides a neural-fuzzy PID control method of a four-rotor aircraft based on repetitive control compensation. The method comprises the following steps that S10) a dynamic model of a four-rotor unmanned aerial vehicle (UAV) is established; S20) neural-fuzzy PID control based on repetitive compensation is carried out; S21) a grid structure in which a neural network generates fuzzy inference (rules) and a PID parameter can be adjusted by itself is designed; and S22) repetitive compensation control is carried out. According to the method provided by the invention, repetitive control based on an internal model principle is embedded into self-adjusting PID closed-loop control in which fuzzy inference is generated on the basis of neural network, neural-fuzzy PID control based on repetitive compensation is formed, the system is still in the closed-loop state, neural-fuzzy PID carries out real-time control adjustment on an output error, a repetitive compensation controller carries out adjustment when the system is in the stable state, output signals can effectively track input signals in the stable state, neural-fuzzy PID adjusts the input signals when interference is relatively high, the signal error is reduced, and the tracking precision of the aircraft system is improved.

The invention relates to a stability analyzing and optimizing method suitable for layering and zoning of an ultra-high voltage electric network, and belongs to the field of safety of the ultra-high voltage electric network. The stability analyzing method comprises an improved short-circuiting current level calculating method and an ANFIS (adaptive neural fuzzy interference system)-based safety domain optimum trend analysis. The invention also provides a stability optimizing method suitable for the layering and zoning of the ultra-high voltage electric network, and the stability optimizing method comprises the following steps of establishing a reactive optimizing model based on a layering and zoning strategy, adopting an improved genetic algorithm, and the like. The stability analyzing and optimizing method has the advantages that the stability of a receiving-end electric network is analyzed and optimized by the improved short-circuiting current level calculating method, the ANFIS-based safety domain optimum trend analysis and a reactive compensation optimizing method based on the improved genetic algorithm; the safe and stable running of the electric network can be ensured, and a quantitative support and decision reference is provided for a power company when the layering and zoning planning of a receiving-end system accessed with the ultra-high voltage is made; and the method is a reliable analyzing and decision-making method, and considerable technical benefits, economic benefits and social benefits can be created.

The invention discloses an error compensating method of a photoelectric encoder. An error compensating process of the method is simpler, so as to realize the aim of improving the compensation precision. Equipment provided by the method comprises the photoelectric encoder, a rotating table and a controller. The error compensating method comprises the following special steps of: obtaining y by subtracting z by x, wherein each angle measurement value of the controller is regarded as a value x to be compensated, and each corner value of the rotating table is regarded as a conventional true value z; according to the x and the z, establishing a knowledge base model of the photoelectric encoder by using a least square method, and outputting z'; according to the x and the y, establishing a neural fuzzy model by using a structure of an improved self-adaption neural fuzzy reasoning system, and outputting ynet(k); adding outputs of the established knowledge base model and neural fuzzy model, so as to obtain z'+ynet(k); and respectively inputting each measurement value of the photoelectric encoder to the established knowledge base model and neural fuzzy model so as to compensate, and outputting the compensated values by the controller.

The invention provides an intelligent electric vehicle adaptive cruise control system and method, and relates to vehicle safety driving assistance control. The system includes an information acquisition module, an operation mode selection module, a control action switching module, a desired torque calculation module, a converter module and an actuator module. A safety distance control strategy anda driving / brake switching strategy are proposed. The neural-fuzzy-based back-stepping sliding-mode adaptive cruise tracking mode control method is adopted, and the nonlinear problem of speed controlof the electric vehicle adaptive cruise control system and the high coupling performance of the system state are solved. The ability of the vehicle tracking a preceding vehicle during adaptive cruisedriving is ensured, the utilization of traffic roads is improved, and the safety and comfort of vehicle driving is improved.

Te invention discloses a battery SOC estimation method applying multiple models parallelly. The method includes following steps: 1), acquiring battery state parameters; 2), normalizing the battery state parameters; 3), respectively substituting the battery state parameters after being normalized into a genetic algorithm optimized BP neural network estimation model, a self-adaptive neural-fuzzy reasoning system model and an OS-ELM neural network model to acquire three SOC estimation results; 4), according to a SOC-OCV relation, acquiring initial estimation voltage values respectively corresponding to the three SOC estimation results acquired in the step 3); 5), respectively calculating difference values between the initial estimation voltage values and voltage, and normalizing the difference values to acquire weighting coefficient of the initial estimation voltage values; 6), calculating a final estimated value of SOC according to the weighting coefficient. By utilizing three importantparameters of a battery, using the three models parallelly, using multiple linear system estimation results to describe a nonlinear system and then weighting, summing and estimating the battery SOC value, estimation accuracy is improved effectively.

Intelligent technique is an effective method to perform the network resource management. A three layer cascade adaptive neural fuzzy inference system (ANFIS) based intelligent controller is proposed for the mobile wireless network to optimize the maximum average throughput, minimum transmit power and interference for multimedia call services. The proposed intelligent controller is designed with a three layer cascade architecture, which mainly contains an ANFIS rate controller (ARC) in the first layer, an ANFIS power controller (APC) in the second layer and an ANFIS interference controller (AIC) in the third layer. The design aim of the proposed three layer cascade ANFIS cognitive engine is maximizing the average throughput of the mobile wireless network, while minimizing the transmit power and interference power.

The invention provides a method for determining the residual electric quantity of a battery, a device, and an automobile, and relates to the technical field of automobile. The method comprise the following steps: choosing an electric quantity determining mode from an established first electric quantity determining mode and an established second electric quantity determining mode, wherein the first electric quantity determining mode and the second electric quantity determining mode are both a five layer structure, which is constructed on the basis of artificial neural network and an adaptive neural fuzzy system of fuzzy logic; inputting the obtained current state information of a battery into the target electric quantity determining model to obtain the residual electric quantity of the battery. The provided method solves the problems that in the prior art, the residual electric quantity is hard to determine under a work condition, the requirements on the current sampling precision are high, and the error of the results is large.

The present invention describes a device incorporating artificial intelligence and corresponding methods for recommending an optimal machinability data selection, especially with machine performance degradation. The device comprises of a first component, which feeds the system with necessary inputs. A second component, which is the main processing unit, acts as an inference engine to predict the outputs. The last component interprets the outputs, conveys the processed outputs to target location and converts them into necessary tasks. The inputs are identified as the machining operations, work piece material, machining tool type, and depth of cut. The input includes machine performance characteristics as well, that is the degradation level of the machine which interrelates with machine vibration and surface finishing. The outputs are the machining parameters, comprising of the optimal cutting speed and feed rate. The inference engine can be established with fuzzy logic, neural network or neural-fuzzy.

The invention relates to an online diagnosis method for a linear control system network. The method comprises the steps that first, a model including two adaptive neural fuzzy inference systems is established, linear control system network state parameters obtained in a fault environment are used as sample data input, a mixed learning algorithm is used for conducting offline training on the two adaptive neural fuzzy inference systems, fuzzy control rules are extracted respectively, then the fuzzy control rules are converted into a loop-up table, a network health degree calculation function and a network fault reason judging function are generated and are added into a controller node of a linear control system, an linear control system online diagnosis algorithm based on the adaptive neural fuzzy inference systems is formed, and online diagnosis is conducted on the network health degree and the fault reason of the linear control system.

The invention discloses a D-S evidence theory-based high-voltage circuit breaker mechanical fault diagnosis method.. The method comprises the steps of establishing a self-adaptive neural fuzzy inference system model; acquiring a sound signal and a vibration signal of the high-voltage circuit breaker; extracting characteristic quantities of the sound signal and the vibration signal; performing classification preprocessing on the characteristic quantity through the adaptive neural fuzzy inference system model to obtain a basic probability distribution function; enabling a D-S evidence theory modified by high conflict evidence to fuse a basic probability distribution function to obtain a mechanical fault diagnosis result.. According to the diagnosis method provided by the invention, the soundand the vibration signal are combined. The D-S evidence theory is used for information fusion to comprehensively diagnose the mechanical failure of the high-voltage circuit breaker, which improves the reliability and accuracy of the state assessment.

A method for predicting a near-ground wind field point domain mapping space along a railway comprises the following steps that S1 a three-dimensional model of a near-ground wind field is set up; S2 an installation position of a wind meter is determined, and under the restraint of the installation position of the wind meter, an arranging scheme of wind measuring points is determined with the least favorable wind acceleration factor and the largest turbulence intensity as targets; S3 a high wind speed numerical simulation model is set up, and a mapping relational expression of the wind speed at the wind measuring points and the wind speed of any one point in the three-dimensional model of the near-ground wind field is provided; S4 according to a predicted wind speed sequence and a mixed algorithm based on empirical mode decomposition and the self-adaptive neural-fuzzy reasoning, prediction of the wind speed along the railway to be predicted is achieved. The method for predicting the near-ground wind field point domain mapping space along the railway can monitor the wind speed of a point along the railway to predict continuous wind speed distribution of a domain, overcome the defect that a train is guided to run only according to real-time wind speeds before and behind the train in the prior art, and improve running safety coefficient of the train in a high wind environment.

The invention presents a self-optimizing deployment scheme and device for indoor small cell. Based on time division multiplexing access (TDMA) resource assignment, an adaptive neural fuzzy inference system (ANFIS) based self-optimizing deployment cascade control (SODCC) device is proposed for indoor small cell to enable easy installation, multi-user (MU)service reliability, optimum throughput, power saving, minimum interference and good cell coverage. The proposed SODCC device is designed with a cascade architecture, which mainly contains an ANFIS based initial power setting controller (IPSC) in the first unit, an ANFIS based channel quality index (CQI) decision controller (CQIDC) in the second unit and an ANFIS based self-optimizing power controller (SOPC) in the third unit. In addition, a protection mechanism of the self-optimizing power controller is included in the SODCC device to prevent the co-channel is interference from the moving users of adjacent cells.

The invention discloses an air gap eccentricity fault diagnosis and classification method of an ANFIS wind power double-fed asynchronous motor, belonging to the field of motor state detection and fault diagnosis. Wind power double-fed asynchronous motor air gap eccentricity faults are divided into several frequent fault types, based on software simulation. A double-fed asynchronous motor model is simulated and several fault types when an air gap eccentricity happens are simulated. The changes of current in a stator winding under different eccentricities of moving and static eccentric faults, a time domain is converted into a spectrogram when the wavelet decomposition of the current is carried out, characteristic frequency bands when different faults happen are extracted, the fault characteristic frequencies corresponding to different types of the air gap eccentricity faults are analyzed, then the wavelet energy of the bands are used as training sample data, an adaptive neural fuzzy inference system for the double-fed asynchronous motor air gap eccentricity faults is constructed, a hybrid learning algorithm is introduced to carry out training, and the air gap eccentricity fault type of the double-fed asynchronous motor is judged. The method has the advantages of high precision and high operability.

The invention relates to an intelligent photovoltaic array fault diagnosis method based on an adaptive neural network, and the method comprises the following steps of S1, collecting photovoltaic electrical characteristic data under various working conditions, and forming original fault data through the sampling and filtering processing; S2, performing data mapping operation on the original fault data to obtain overall fault feature data; S3, performing feature dimension reduction compression on the overall fault feature data to three dimensions by using an LDA algorithm to obtain new feature data; S4, dividing the new feature data into a test set and a training set by adopting K-fold cross inspection, and setting the number of membership functions and the types of the membership functions;S5, generating an initial fuzzy inference system; S6, constructing an adaptive neural network fuzzy inference system model; S7, judging whether the photovoltaic array system is in a fault state or not. The technology of the invention can effectively diagnose and classify the photovoltaic array in the fault, and compared with other machine learning algorithms, the classification precision is highand the result is accurate.

The invention provides a method and an apparatus for dynamically and safely determining corrosive defects inside a conveying pipeline. The method comprises the following steps: acquiring a corrosion expansion rate of a conventional metal material for a conveying pipeline under different CO2 partial pressure, H2S partial pressure, chloride contents and different stress levels; establishing a mathematic model of the corrosion expansion rate by utilizing a self-adaptive neural-fuzzy inference system according to the acquired corrosion expansion rate; acquiring the CO2 partial pressure, a H2S content, the chloride content and the stress level of a pipe section at which a corroded portion of the pipeline is located, and predicting the corrosion expansion rate according to the established mathematic model; and establishing a pipeline limit-state equation, and acquiring a safety state of the pipeline. The invention also provides the apparatus for dynamically and safely determining the corrosion defects inside the conveying pipeline. By adopting the method and the apparatus for dynamically and safely determining the corrosion defects inside the conveying pipeline, the problem that the safety state of the oil-gas conveying pipeline containing the corrosion defects is difficult to determine dynamically and accurately can be solved.

The invention relates to a neural fuzzy stability control system of an electric automobile charger and a control method. Compared with a traditional full-bridge type DC-DC conversion circuit, a phase-shifted full-bridge ZVS DC-DC conversion technique adds a resonance inductor and a resonance capacitor element to a circuit, resonance of the resonance inductor and a capacitor is used, four power switch tubes of a full-bridge circuit work in a zero-voltage turn-on and turn-off soft switching state, so that the problem of high-frequency switching power consumption is solved, and electric energy conversion efficiency is improved. A neural fuzzy PID control technique combines neural network control and fuzzy PID control, robustness is good according to fluctuation of circuit parameters, and meanwhile, adaption to the characteristic of object time varying nonlinearity can be achieved. By the adoption of the two techniques, the overshoot and fluctuation of circuit output voltage can be reduced, the circuit response rate is increased, external disturbance is restrained, and circuit stability is improved.

A track circuit fault monitoring platform and a monitoring method based on a frequency shift signal are disclosed. The monitoring platform is designed and completed based on LabVIEW and carries out real-time processing on front end detection data. The monitoring method is characterized in that a portable frequency shift signal detector detects a key parameter of a frequency shift signal in a steelrail and uploads data to a cloud server; and the monitoring system downloads time frequency information of the frequency shift signal from the cloud server in real time, extracts a time domain of thefrequency shift signal and a frequency domain characteristic value and takes as fault characteristic quantities, an adaptive neural fuzzy inference system is used to carry out fault classification, and finally early warning information of a ZPW-2000A non-insulation track circuit is given. In the invention, the frequency shift signal can be effectively used to monitor a track circuit fault, and the platform and the method can be used as a supplement of railway electric service section outdoor monitoring.

Disclosed is an intelligent control method for a semi-active suspension system of an automobile. The method comprises the steps that a fuzzy controller is improved, a neural fuzzy inference system based on a Mamdani model is designed to serve as a controller, in combination with back propagation (BP) learning rules, fuzzy control rules are obtained through training, and by adopting a Simulink S-function, a corresponding fuzzy neural network is constructed to control a simulation model of the semi-active suspension system of the automobile. By analyzing a simulation test result of the model, itis shown that through the utilization of the model, the comprehensive performance of the suspension system of the automotive can be significantly improved. Under same experimental conditions, a simulation test is uniformly carried out on a passive suspension and a semi-active suspension system model based on fuzzy control and Mamdani fuzzy neural network control in the Simulink environment. Corresponding root mean square values of suspension performance evaluation indexes including the vertical acceleration of an automobile body, the dynamic travel of the suspension and the dynamic load of atire serve as output of a simulation module. By means of the intelligent control method, the performance of the semi-active suspension system can be effectively improved.

The invention provides a noise suppression Capon active target DOA estimation method based on time reversal, which is different from most passive target DOA estimation. The method includes steps: an active target signal is received by employing a passive antenna array, namely the antenna array does not actively emit a signal detection space target but passively receives a signal emitted by an active target; when the antenna array receives the signal emitted by the active target, the signal is converted to receiving wave matrix forms for different channels, then time reversal and value back-transmission are performed, and a neural fuzzy reasoning system is adapted for noise suppression; and the noise suppression Capon active target DOA estimation method based on a TR is proposed, a space-time matching focusing characteristic of the TR and a specific direction optimization characteristic of a Capon algorithm are combined, and the estimation precision of a DOA is further improved. According to the method, multipath regarded as clutters is fully adopted, adequate receiving wave useful information is provided, and a result shows that an active target DOA spectrum obtained by employing the method has low sidelobe, high resolution and high precision, and the lower limits of the RMSE and the CRLB are lower.

The invention designs a magneto-rheological damper hybrid modeling method. The method mainly comprises the four steps of: performing mechanical property experiment on a magneto-rheological damper, analyzing the power indication and speed characteristics, and obtaining the original experiment data of the piston displacement, the speed, the current and the damping force; performing numerical filtering and normalization processing on the original experimental data, and creating an input and output sample set; modeling the magneto-rheological damper by adopting a self-adaptive neural fuzzy reasoning system, and introducing a subtractive clustering technology to construct a system rule base; and optimizing the clustering parameters by using a genetic algorithm, establishing a magneto-rheological damper model, and analyzing the model precision, the hysteresis characteristic and the like. According to the method, a subtractive clustering technology is introduced to construct a rule base of a magneto-rheological damper model, a genetic algorithm is utilized to optimize clustering parameters, and the optimal fuzzy rule quantity and structure for depicting the system behavior are obtained, so that the modeling accuracy is effectively improved, the hysteresis characteristic and the low-speed region behavior of the system are better described, and the development and application of a semi-active control technology are promoted.

The invention discloses a modeling method of an optimal power flow model of a receiving end power grid security domain. The method includes a first step: constructing a system security domain through analysis of stability of a power flow equation, namely stability of quiescent voltage, phase angles and oscillation frequency of a system, and N-1 operation standards, wherein N loads of the system can be converted into M different sets composed in the load direction according to given power generation dispatching standards to generate a critical load matrix for approximate treatment of the security domain; a second step: constructing a dynamic security constraint optimal power flow model; a third step: constructing a self-adaptive neural fuzzy inference system; and a fourth step: training the self-adaptive fuzzy inference system and constructing the optimal power flow model. The optimal power flow model well expounds the operation standards of current electric power system dispatching. The optimal power flow security domain approximation technology based on the self-adaptive fuzzy inference system can be applied to dispatching optimization between receiving end power grid areas after distribution of hierarchy and areas.

The invention discloses a method for compensating a temperature error of a micro-electromechanical inertial measurement unit. The method comprises the following steps: designing a full-temperature experiment and collecting full-temperature output data; according to the output data of the full-temperature experiment, analyzing the temperature output characteristics of the MIMU, and selecting temperature variables to establish a temperature error model of an MEMS accelerometer and an MEMS gyroscope; designing an adaptive neural fuzzy inference system; respectively inputting the temperature errors of the MEMS accelerometer and the MEMS gyroscope as training samples into the adaptive neural fuzzy inference system; carrying out trainingby using an adaptive neural network to obtain fuzzy parameters and an acquiringan optimal network model; and calculating prediction outputs of the temperature errors according to the optimal network model, and compensating the full-temperature output of the MIMU by applying a network prediction result. According to the method, a learning mechanism of the neural network enters fuzzy reasoning, so that the temperature error modeling precision is improved, the temperature drift of the MEMS gyroscope is accurately compensated, and the precision of the MIMU in a full-temperature range is realized.

The invention discloses a control method for steer-by-wire automobile active front-wheel steering. The control method sequentially comprises the following steps that (1) a self-adaptive neural fuzzy inference model is built; (2) the initial weight value and the threshold value of the self-adaptive neural fuzzy inference model are optimized through the genetic algorithm; (3) the self-adaptive neural fuzzy inference model optimized through the genetic algorithm is applied to an active steering controller; (4) a full automobile model is built; (5) the steering wheel rotating angle and the actual automobile speed serve as input of the full automobile model, and the deviation of the ideal yaw velocity output by the full automobile model and the actual automobile yaw velocity serves as input of the active steering controller; (6) the active steering controller outputs the additional front wheel rotating angle; and (7) the additional front wheel rotating angle and the steering wheel rotating angle are overlapped and fed back to an automobile steering actuator. By means of the control method for a steer-by-wire automobile active front-wheel steering control system, the automobile stability during automobile steering is improved, and the safety of a driver in the automobile running process is ensured.

The invention discloses an ANFIS-based electric vehicle permanent magnet synchronous motor fault classification method. The method comprises the following steps: faults are classified, and a training sample set is built through acquiring various fault data; and an adaptive neural fuzzy inference system is built, winding current in the fault data for various fault types in the electric vehicle permanent magnet synchronous motor serves as an input, one output is given for each fault type, a membership function for the input and the output is selected, system training target errors are set, a hybrid learning algorithm is used for training parameters of the membership function, and thus, input membership function parameters and output membership function parameters in the adaptive neural fuzzy inference system are thus determined. Through diagnosing the faults of the permanent magnet synchronous motor, experimental data are obtained, the experimental data are inputted to the adaptive neural fuzzy inference system, a diagnosis result is obtained, and a fault type is determined according to the diagnosis result, and thus, fault classification is completed. Strong-operability, high-efficiency, economic and high-accuracy diagnosis is realized.

The invention provides a haze prediction method and device and belongs to the meteorological prediction technical field. The method includes the following steps that: monitoring points in a monitoring area are determined; a plurality of effective observation values are selected according to the data information of the monitoring points in the monitoring area so as to be adopted as feature selections for haze prediction, and neural fuzzy models are built for the feature selections, wherein the data information includes first data information and second data information, and the neural fuzzy models include an MLR model, an ANN model and an NF model; subordinating degree functions are selected according to the first data information and the second data information of each monitoring point, and the first data information and the second data information are multiplied, so that third data information can be obtained, normalized confidence can be calculated based on the third data information; and a fuzzy result is calculated according to a fuzzy rule and the normalized confidence result of each monitoring point, and a haze prediction result is obtained according to the fuzzy result. With the haze prediction method and device of the invention adopted, the real-time performance, validity and reliability of haze prediction can be effectively improved.

The invention discloses a method for predicting ground surface settlement caused by shield construction based on a neural fuzzy inference system. The method comprises the steps of selecting main influence factors of ground surface settlement caused by shield construction, forming an effective data set by the main influence factors, calculating the effective data set by means of a ground surface settlement prediction model built by the neural fuzzy inference system, and therefore obtaining a prediction result and performing model verification according to the prediction result. Accordingly, the method for predicting ground surface settlement has the advantage of being high in prediction precision, and is easy to use, small in calculated quantity, high in generalization, thereby having high practicability, and the novel technical scheme can be provided for predicting ground surface settlement caused by shield construction.

The invention relates to an image edge detection method based on self-adaptive neural fuzzy inference systems. With regard to an image polluted by spiced salt pulse noises, the method has the advantages that a network containing four self-adaptive neural fuzzy inference systems and a post-processing module is constructed; before the network is used for carrying out edge detection on a noise image, a training image is artificially constructed and a mixed learning algorithm is used for independently training the four self-adaptive neural fuzzy inference systems to determine parameters in the systems; after the four self-adaptive neural fuzzy inference systems are trained, one network can be formed by the four self-adaptive neural fuzzy inference systems and one post-processing module to carry out the edge detection on a testing image. The image edge detection method disclosed by the invention has the characteristics that even if the testing image is polluted by the noises, the method can effectively extract edge information in the image and does not need to carry out image filtering pre-processing process.