101 results about "Nonlinear prediction" patented technology

The present invention relates to methods and devices for noninvasive nonlinear prediction of ictal onset in patients afflicted by neurological disease. In particular, the present invention provides methods and devices for noninvasive nonlinear prediction of seizures in patients afflicted with epilepsy. The devices and methods preferably being based on analysis of two or more electroencephalogram (EEG) recordings, one set of recordings taken from an electrode close to the region of ictal onset, and a second or more set of recordings (e.g., concurrent readings) taken from a region remote from the region of ictal onset.

The invention discloses a combustion optimization control method for a boiler. The combustion optimization control method is characterized by comprising the following steps: sampling a combustion nonlinear system of the boiler to obtain input/output data at the current moment; training the real-time sampled input/output data by an online incremental learning fuzzy neural network, building an online incremental learning predicting model of the combustion nonlinear system of the boiler; performing a nonlinear prediction control algorithm on the online incremental learning predicting model for realizing the optimization and the control of the combustion process of the boiler. According to the combustion optimization control method for the power station boiler of the online incremental learning fuzzy neural network, the nonlinear optimization problem in the predication control algorithm is solved by utilizing a particle swarm optimization algorithm through the online identification of the boiler combustion optimization model; the real-time optimization and control of the boiler combustion process are realized.

The invention discloses a method for generating a variable parameter chaotic signal, and a chaotic secure communication system. The method comprises the following steps: firstly, chaotic mapping is used to generate the chaotic signal; then a parameter set of the chaotic mapping is determined in advance, so as to process any pseudorandom signal, and enable the state of the pseudorandom signal and the elements of the parameter set to be in one-to-one correspondence; and a corresponding parameter is chosen to generate the variable parameter chaotic signal by chaotic mapping according to the state of the pseudorandom signal. The system comprises a chaotic signal generating module, an analog-to-digital conversion module, a parameter transformation module, a coding module, a multi-path choice switch, an encryption/decryption module, an encrypted message data cache, a plaintext data cache and a control module. The invention can increase the complexity of the output of the chaotic signal or a chaotic sequence, and effectively resist the analysis of the nonlinear prediction technology based on phase-space reconstruction. The output digital chaotic sequence can be taken as a pseudorandom number sequence or key stream to encrypt data; the output continuous chaotic signals can be used for designing a secret communication system based on chaos synchronization .

The invention discloses a method for performing nonlinear prediction on coke quality on the basis of cohesiveness and coal-rock indexes of single coal, which provides important technical guarantee for improving the stability and quality of coke produced by coke making enterprises. The method comprises the following steps of: establishing an information database storing qualities and coke quality indexes of the single coal for coking, and inputting the cohesiveness indexes and the coal-rock indexes of the single coal for coking into the coal information database; establishing a coal quality prediction model, and predicting quality indexes of matched coal by virtue of a clustering and support vector machine; defining the quality indexes of the matched coal, and predicting quality indexes of the coke according to the quality indexes of the single coal; and establishing a model for predicting the crushing strength and abrasive resistance of the coke. The invention further provides a coke quality prediction system with optimized coal blending of coal-rock, which aims to stabilize and improve the coke quality and reduce the coal blending cost, can form a prediction model with multiple parameters and high accuracy for the coal quality indexes and the coke quality indexes, and simultaneously, has a real-time updating function or a manual intervention function.

The invention relates to a lysine fermentation process feeding prediction control system and method based on a fuzzy neural network. The control method comprises the following steps: firstly using batch data to train the fuzzy neural network and establish a nonlinear prediction model of lysine fermentation process, secondly measuring the input/output information of model utilization history and future batch data to predict the future output of the lysine fermentation process, using the model to output error and perform feedback compensation so as to obtain closed loop output, finally comparing closed loop output and reference input trajectory, utilizing quadratic form performance index to perform rolling optimization, and calculating to obtain feeding control quantity which is needed to add in the system currently. The system comprises a site intelligent detecting instrument and peristaltic pump which are directly connected with a fermenter and an intelligent controller, wherein feeding prediction control algorithm is embedded in the system intelligently. The system and method of the invention can adapt to the dynamic characteristic of the lysine fermentation process and the coupling between strong open loops, thus obtaining good control effect.

The invention is suitable for the technical field of power supply, and provides a method and device for predicting carbon emission, a terminal and a computer readable storage medium, and the method for carbon emission prediction comprises the steps: obtaining historical carbon emission data and enterprise information data within a set time, building an autoregressive moving average (ARMA) model according to the historical carbon emission data and the enterprise information data, and obtaining a linear prediction model of the carbon emission; calculating a residual sequence based on the historical carbon emission data and a prediction result of the linear prediction model; constructing a support vector machine (SVM) according to the residual sequence and the enterprise information data, and obtaining a nonlinear prediction model of the carbon emission; and combining the linear prediction model and the nonlinear prediction model to obtain a target prediction model of the carbon emission. The invention can achieve the prediction of the carbon emission in regional industrial planning and construction, provides a reference basis for the formulation of a power supply strategy, and improves the power supply efficiency.

The invention discloses a rapid detection method of the milk protein content based on a dielectric spectrum technique. According to the rapid detection method, dielectric spectra (a relative dielectric constant spectrum and a dielectric loss factor spectrum) of a batch of milk samples within a radio frequency/microwave range are obtained by utilizing a network analyzer (or an impedance analyzer) and a coaxial probe, characteristic dielectric variables which express the milk protein content are extracted from the dielectric spectra, linear or nonlinear prediction models which are used for predicting the milk protein content are established on the basis of the dielectric spectra or the extracted characteristic dielectric variables within the whole frequency range, the models are inspected, and the model of which the root-mean-square error is the minimum is used as the optimum model for predicting the milk protein content. Data of the dielectric spectra under unknown characteristic frequency of the milk samples are substituted into the optimum model, so that the protein content of the samples can be calculated. The rapid detection method has the advantages of rapidity in detection, convenience, high precision, simplicity in operation and capability of real-time online detection.

The invention discloses a self-detection method for positions of rotors of switched reluctance motors. The self-detection method includes building prediction models of relevant vector machines and acquiring parameters of particle swarm optimization models; estimating position angles of the rotors by the aid of the nonlinear prediction models for the positions of the rotors; performing self-detection on the positions of the rotors of the motors. By the aid of the self-detection method, shortcomings of long decision-making time, poor instantaneity and the like due to complicated models in the traditional intelligent detection method can be overcome, and the self-detection method has the advantages of good generalization ability, short online computation time, high prediction precision and the like, and is applicable to performing quick self-detection on the positions of the rotors of the switched reluctance motors in full-speed running ranges.

The invention discloses a bidirectional grid connected inverter, including a power supply Udc, a capacitor C and three inversion circuits. The capacitor C and the three inversion circuits are connected in parallel with the power supply Udc. The inversion circuits are provided with three paths of inductors L, capacitors R, inductors L1, resistors R1 and alternating current power supplies connected in series therebtween, and a capacitor C1 is connected in parallel between each capacitor C1 and each inductor L1. The alternating current power supplies comprise Ua, Ub and Uc. According to the invention, as the traditional PID control strategy exerts obvious limitation on the control of voltage, current and power of an inverter, an advanced intelligent control theory is adopted for timely measuring and controlling a large-power and high-voltage bidirectional grid connected inverter system; and a nonlinear prediction model which is achieved by applying a RBF neural network has advantages of fast convergence rate, prevention of convergence at a local extreme point and concise network structure, thereby further improving the dynamic response characteristic of the system.

The invention relates to the technical field of shield underneath pass existing tunnel construction multi-objective optimization and discloses a shield underneath pass existing tunnel construction multi-objective optimization method based on GA-LSSVM and NSGA-II. The method mainly comprises the following steps of S1, collecting data of horizontal displacement and settlement displacement of an existing tunnel arch bottom based on shield construction parameters; S2, a GA improved least square support vector machine (GALSSVM) being adopted to establish a high-precision prediction model of horizontal displacement and settlement displacement of the arch bottom of the existing tunnel, and two regression prediction functions being obtained; and S3, taking the two nonlinear prediction functions asfitness functions, combining the application constraint conditions of the influence factors, and performing multi-objective optimization by utilizing NSGA-II to obtain an optimal mix proportion. By means of the established GA-LSSVM and NSGA-II model, high-precision prediction of arch bottom horizontal displacement and settlement displacement is achieved, and multi-target intelligent optimizationof shield construction parameters is also achieved.

The invention, which relates to the technical field of cut tobacco drying process moisture control, discloses a cut tobacco drying process moisture prediction control method and system based on a recurrent neural network. The method comprises the steps: step A, collecting related data of a cut tobacco drying process; step B, automatically identifying acquired trademark information to obtain control parameters; step C, judging the related data, and establishing a nonlinear prediction control model; step D, converting a nonlinear prediction model into a nonlinear prediction control model based on a recurrent neural network, and updating the weight of the recurrent neural network to obtain an outlet moisture content prediction value; and step E, constructing a performance index J, and acquiring an optimal moisture removal air door opening degree of the performance index J. According to the method, the nonlinear prediction control model is improved, the neural network training speed and stability are improved, and the stability of the outlet moisture content is improved.

The invention provides a nonlinear prediction control system for an energy-saving air separation process, which comprises a field intelligent instrument and a DCS system which are directly connected with an air separation tower. The DCS system comprises a memory device, a control station and an upper computer, wherein the intelligent instrument is connected with the memory device, the control station and the upper computer; the upper computer comprises a nonlinear prediction controller which has a function of optimizing and solving a control law output operation variable value; and the nonlinear prediction controller comprises a component inferring module, a model parameter self-adaptation correction module and a control law rolling optimizing solution module. The invention also provides a nonlinear prediction control method for the energy-saving air separation process. The invention provides the nonlinear prediction control system and the nonlinear prediction control method for the energy-saving air separation process, which can effectively achieve highly accurate tracking control effect, have a fast online solution speed and greatly improve the work efficiency.

The invention belongs to the technical field of cement fineness prediction, in particular to a cement particle size distribution prediction method based on random distribution. The method includes thesteps of obtaining the feed quantity of mill, working current of mill, temperature of inlet and outlet of mill, pressure difference of mill, current of circulating lifting machine, rotational speed of separator, current of circulating fan and probability distribution of cement particle size at corresponding time, and storing all parameter signals as historical data set; S200 establishing the basis function expression model of probability distribution density function of cement particle size; S300 screening out abnormal data, and assigning sample weights according to classification to form newdata samples; S400 establishing the nonlinear prediction model between the input variables and the previous n-1 weight vectors to predict the cement particle size distribution at the next time; S500updating the basis function of the cement particle size probability distribution density function to represent the model parameters through the output error value of the model. The method of the invention can detect the cement particle size distribution in the cement grinding system in real time.

The invention belongs to the field of coal quality detection and discloses a coal quality online detection and analysis method based on regression analysis. Graphic features of various areas in coal are automatically extracted by regression analysis algorithm of machine learning for image characteristics extraction; and a coal quality online detection result is provided based on the extracted image characteristics by machine learning, and the detection result is displayed through a display module. the method of the invention comprises the following specific steps: a coal scanned image is acquired through data acquisition equipment with built-in near-infrared detection module and laser detection module; the coal scanned image is pretreated by the use of a pretreatment module, and the coal scanned image is divided into multiple blocks. Coal quality data information detection is carried out through the near-infrared detection module and the laser detection module, and the detection data is accurate. By component network for nonlinear prediction, the method of the invention has advantages of strong adaptability, good fault tolerance and the like. All elements analysis of coal quality can be realized, and measurement accuracy also can be enhanced.

The invention discloses a pneumatic-thermal collaborative optimization method for a scallop hole air film cooling structure of a turbine blade. According to the method, a pneumatic-thermal feature agent model of the scallop hole air film cooling structure of the turbine blade is established based on a radial basis neural network, and a particle swarm optimization algorithm is introduced to realize pneumatic-thermal collaborative optimization of the scallop hole air film cooling structure of the turbine blade. The method overcomes the defect that a traditional air film cooling structure optimization method needs to depend on a large quantity of samples. The method has a nonlinear prediction ability, high prediction precision, a strong memorizing ability, high robustness and a good global approximation ability. Moreover, weights of pneumatic performance and thermal performance can be adjusted according to actual demands to design an optimal scallop hole air film cooling structure with both the pneumatic performance and the thermal performance.

The invention relates to a fuel cell vehicle energy management method based on nonlinear prediction model control, which comprises the following steps of 1) acquiring the working condition information of vehicle driving in real time, and judging the working condition state of vehicle driving in the current time, 2) inputting the state parameters of the current working condition in the step 1) into the established prediction model, and determining the running state of the vehicle in a future prediction time domain, 3) according to the running state of the vehicle in the future prediction time domain in the step 2), optimizing a control sequence in a limited time domain of the vehicle by utilizing a dynamic programming algorithm according to the requirements of economy and durability of a power supply system, and 4) applying the first element of the optimal control sequence of the finite time domain obtained in the step 3) to the vehicle, and repeating from the step 1) at the next moment. Compared with the prior art, the method has the advantages of reliable performance, comprehensive better economy and durability of the power supply system, strong practicability and the like.

The invention discloses an electric automobile compound energy management method based on a rule and nonlinear prediction control. By the adoption of the method, energy management is conducted according to the power requirement of a vehicle at every moment and the SOCs of a lithium battery and a super capacitor. In a nonlinear prediction control strategy, a controller will predict the speed in a certain period in the future, the speed is converted into power through a vehicle running speed and power model, the output current of the lithium battery is optimized with the minimum power consumption as an index through a quadratic programming effective set method, and power distribution of the lithium battery and the super capacitor is finished. By the adoption of the method, on the basis thatthe required power is met, system energy loss can be reduced, use of the lithium battery is reduced, the service life of the lithium battery is prolonged, and the efficiency of a hybrid power system is improved.

The invention relates to a nonlinear unstable-combustion prediction method and device. The nonlinear unstable-combustion prediction method comprises the steps of selecting a pulsating pressure time sequence with preset length in a combustion chamber; utilizing an autocorrelation function method to calculate time delay of the pulsating pressure time sequence; adopting a G-P algorithm to calculate embedded dimension of the pulsating pressure time sequence according to a Taken theorem; reconstructing a phase space according to the time delay and the embedded dimension of the pulsating pressure time sequence; performing singular value decomposition on a matrix of the reconstructed phase space and extracting a maximum Lyapunov index of the pulsating pressure time sequence; establishing a nonlinear prediction model of the pulsating pressure time sequence based on the maximum Lyapunov index and performing nonlinear prediction on development of the pulsating pressure time sequence in the combustion chamber according to the nonlinear prediction model. By means of the nonlinear unstable-combustion prediction method and device, a nonlinearity characteristic of the pulsating pressure time sequence in the combustion chamber can be effectively caught to serve as an important basis for combustion adjustment.

The invention relates to the technical field of application of servo press, and particularly discloses a full-closed-loop nonlinear prediction control method and system for a servo press. According tothe full-closed-loop nonlinear prediction control method and system, a sliding block of the servo press is used as a direct control object, and the full-closed-loop control covers the position control of a motor and the position control of the sliding block; due to the fact that the precision of a grating ruler position sensor, installed on the sliding block, is quite high, factors such as transmission system errors and motor control errors of the servo press can be fully considered, the workload in the calibration process is greatly reduced, and the direct full-closed-loop control method ofthe press can be realized. According to the full-closed-loop nonlinear prediction control method and system, the sliding block position of the servo press can be directly controlled, and the control is performed by applying a generalized predictive control algorithm strategy to the servo motor so that the position of the sliding block can be adjusted on line in real time.

The invention discloses a pile top displacement nonlinear prediction method considering sliding bed rock mass structure characteristics. The method aims at solving the problems that current research on sliding bed rock mass structure characteristics is few, and only stays on a continuous medium-based single-factor sensitivity analysis level; according to the method, multiple factors of sliding bedrock mass structure characteristics are considered, a particle swarm optimization algorithm and a support vector regression machine method are adopted, an optimal prediction model of pile top displacement is established, and when a sliding bed is of a composite layered rock mass structure, the precision of sliding-resistant pile displacement prediction is effectively improved.

The technology disclosed relates to differential analysis of sets of time series pairs. In particular, it relates to building estimators of magnitude of difference between two time series. After the basic estimators are built, they are combined into ensemble estimators using linear or nonlinear prediction models to improve their accuracy. In one application, the ensemble is used for estimating the magnitudes of difference over sets of metric pairs observed from distributed applications and systems running over a computer network. The metric pairs are then ranked in decreasing order of magnitude of difference to guide an operator in prioritizing his root cause analysis of faults, thereby reducing the time-to-resolution of problems.