114 results about "Slow convergence" patented technology

The present application relates to tracking a position of a device, e.g. for detecting slow and rapid earth deformation, by making use of a recursive filter having the filter characteristic adapted to a detected type of motion. If the motion of the position tracking device is rapid, the filter characteristic is set such that the rapid motion can be tracked with the necessary speed. On the other hand, if the motion is slow, e.g. during times of a normal tectonic drift, the filter characteristic is set such that the motion is slowly tracked with the advantage of efficient noise reduction, i.e. noise in the input signal is effectively barred and does not pass through the filter to the output signal. Thus, in times of rapid motion the convergence speed of the filter output signal to the input signal is set high for fast convergence and in times of slow motion the convergence speed of the filter output signal to the input signal is set low for a slow convergence. The filter may be a Kalman filter.

The invention protects a cross-updated active noise control system based on a novel algorithm for online identification of a secondary channel. The noise control system comprises 6 modules of noise signal filtering, momentum FxLMS algorithm, white noise generator, secondary channel modeling, main channel path and third adaptive filter update module. The object of the noise control system is to solve the problem that an active noise cancellation (ANC) system has a slow convergence speed and a small noise reduction amount in indoor noise-cancellation applications. The innovation lies in the factthat the unevenness of the power spectral density of a noise signal greatly affects the convergence speed of a control filter and a modeling filter in the indoor noise-cancellation applications. Themomentum FxLMS algorithm is proposed to update the weight of the control filter, and a variable step size LMS algorithm is used to update the weight of the modeling filter. The third adaptive filter using the Newton LMS algorithm is used to eliminate an error signal and the signal related to the reference input signal, and improves the modeling accuracy of the modeling filter and the convergence speed of the entire ANC system.

The invention discloses a hyperspectral image classification method based on a cooperative generation antagonistic network and a space spectrum combination. The method comprises the following steps: inputting a hyperspectral image; obtaining a sample set; generating training samples and test samples; building a multi-scale discriminator; constructing cooperative relationship; building a cooperative generation countermeasure network; the initial value of the training sample being generated by the multi-scale discriminator; the generator generating samples; a multi-scale discriminator perfromingclassification; constructing the loss function of generator and multiscale discriminator; alternatively training the generator and the multi-scale discriminator; hyperspectral images being classified. The invention utilizes the established cooperation to generate the antagonistic network, extracts the spatial spectrum combined characteristics of pixels, generates more realistic samples, increasesthe number of samples, alleviates the problems of over-fitting the network and slow convergence speed of the network, and improves the accuracy of the hyperspectral image classification.

The invention relates to a reliability analysis method of an electric power system, specifically a wind power grid-connection system operational reliability evaluation method considering uncertainty factors, and solves the problems of incomplete consideration, long time consumption, slow convergence, resource occupation and incapability of realizing real-time prediction and evaluation of the existing method. A scheme comprises the processes as follows: I, uncertainty factor modeling: A, load comprehensive uncertainty; B, a conventional power generator comprehensive time varying operation model; C, comprehensive time varying operation modeling; D, wind power plant operational reliability modeling; and II, a Markov chain model: E, a Markov process and a Markov chain; F, a wind power Markov chain model; and G a wind power reliability evaluation index. The beneficial effects are achieved as follows: by performing quantitative analysis and comparison on the degree of influence of load fluctuation, element state and wind power output on the operational state and reliability level, the reliability level at the current moment can be evaluated, and the state probability at the future momentcan be quickly predicted, so that planning, overhaul, operation optimization and dispatching of the electric power system can be guided, thereby overcoming the shortcomings of the conventional evaluation method.

A single-channel color image encryption method based on chaos and phase retrieval process comprises a color image chaos scrambling and decomposition step, a pure phase extraction and modulation step and a chaos diffusion step. Combining fractional Fourier transform and paired Logistic chaotic mapping, an encrypted image has nonlinearity and randomness, security and convergence speed of an encryption system are raised, and the problem of slow convergence speed existing in a present encryption method is solved. Meanwhile, complexity of the color image encryption system is reduced, the system is simple, and operation is convenient.

The invention brings forward a frequency domain three-dimensional irregular earthquake data reconstruction method. The method is characterized in that first of all, three-dimensional earthquake data in a time domain is converted to a frequency domain by use of Fourier transform, and then, a projection onto convex set (POCS) algorithm is employed and curvelet transform capable of describing localized features of the earthquake data is introduced; and in an iteration process, a new threshold parameter attenuating according to an index rule is brought forward, and each frequency slice is individually reconstructed by use of a soft threshold operator, such that the iteration frequency is reduced, the reconstruction precision is improved, and the purpose of reconstructing the three-dimensional earthquake data is realized. According to the invention, the new threshold parameter attenuating according to the index rule is brought forward and each frequency slice is reconstructed in individually by use of the soft threshold operator, such that the disadvantage of quite slow convergence speed of a conventional threshold parameter is overcome, the calculation complexity of an algorithm is reduced, the calculation efficiency is substantially improved, and the operation time is reduced.

The invention provides a welding spot defect identifying method which comprises a step of performing characteristic extraction on the acquired welding spot image information and a step of identifying the welding spot defect according to the ways of identifying, fuzzy inference, neutral network according to the characteristics and the like, and specifically comprises the following steps of: 1) obtaining a sample for training an artificial neural network according to the principle of an orthogonal test based on the welding spot shape theory; 2) training the artificial neural network by use of an improved neural network algorithm to obtain a network for predicting the possibility of various defects of the welding spots; and 3) performing image processing on the actual welding spot, extracting the shape quality characteristic as the input of the trained artificial neural network, and performing forward calculation by use of the trained network to realize identification of the welding spot defect. In the invention, as the BP neural network is improved and the genetic algorithm is introduced into the neural network algorithm training, the defects of slow convergence, easy trap in local optimal solution and the like of the neural network are solved, and the network performance is improved to some degree so as to realize defect identification of complicated welding spots.

The invention discloses a hybrid spread spectrum communication system based on low density parity code (LDPC) encoding and a method thereof. The problems of leakage capture, high leakage capture probability, slow convergence of a phase-locked loop and low decoding performance of the existing synchronous capture method are mainly solved. The system comprises a transmitting part and a receiving part, wherein the transmitting part is used for carrying out pretreatment on original information; the pretreatment data are subjected to frequency hopping treatment and transmitted after being modulated and filtered by a base band; the receiving part is used for obtaining a synchronization point position and a frequency offset value in the received signal through synchronous capture, and intercepting useful data to carry out de-hopping according to the synchronization point position; the base band demodulates and filters; the filtering data are subjected to frequency offset compensation according to the frequency offset value; the data after frequency offset compensation are subjected to dispreading and frequency offset rectification; the frequency offset rectification data are decoded and the decoding result is output. The hybrid spread spectrum communication system has the advantages of being strong in burst error resisting ability, high in real-time synchronization capture speed, high in capture accuracy, high in convergence speed of the phase-locked loop, and good in decoding performance, and can be applied to reliable communication under the condition of very low powder spectral density.

The invention discloses a calibrating method for an electronic compass. The calibrating method is used for promoting measuring accuracy of the electronic compass based on a self-adaption differential evolution algorithm and a Fourier neural network theory and is especially suitable for an orienting system with low cost and higher precision. The calibrating method comprises the following steps: utilizing a Fourier neural network to perform error modeling on the electronic compass and utilizing an improved self-adaption differential evolution algorithm to optimize a weight value of the Fourier neural network, so as to acquire an accurate error model to compensate a measuring value of the electronic compass. The error model which is established according to the calibrating method is capable of realizing accurate mapping of a sample space and has a higher nonlinear approaching capability. According to the calibrating method, a minimum local part is avoided, the defects of over-slow convergence rate and oscillation of the neural network are overcome and the influence of an outside magnetic field on the electronic compass is efficiently compensated, thereby greatly promoting the measuring accuracy of the electronic compass.

Interfusion between radar image and infrared imagery in pixel level is carried out for complicated city zone by using principal component analysis method. Through combination of back propagation artificial neural network (BP-ANN) and genetic algorithm (GA), BP-ANN / GA hybrid algorithm is built up. By using ability of global optimization of GA algorithm, the method optimizes initial weight value of BP-ANN, overcomes disadvantages of slow convergence rate of traditional BP-ANN, and easy of getting into local minimum so as to increase speed for automatic sorting remote sensing images from multiple sources and accomplish sort for characters of ground objects.

The invention relates to a plane track planning method of an unmanned underwater vehicle (UUV) formation, and belongs to the technical fields of unmanned underwater vehicles and track planning. The method includes: constructing a mission space model according to an environment information database or environment information of sensor detection; using an improved and fused algorithm of an APF algorithm and an RRT algorithm for global track planning on the basis of static environment information; and using global planning path points as sub-target-points, and carrying out local planning in sections on the basis of dynamic environment information. According to the method, the unreachable-target and local-optimization problems of APF are solved through a certain improvement scheme in a globalplanning algorithm, the three major characteristics of RRT non-determination, non-optimization and slow convergence are mitigated, deterministic and stochastic track planning methods are combined, static planning and dynamic planning are combined, location planning and speed planning are combined, a track reduction algorithm achieves an effect of further optimizing a track, and meeting of missionrequirements of safety, optimality and real-time performance of track planning of multiple unmanned underwater vehicles is guaranteed.

The invention relates to a channel estimation method used in an amplitude limiting OFDM system. In the method, pilot frequency information and history channel information are comprehensively used so as to acquire current channel response and improve a convergence speed of an iterative algorithm. Firstly, amplitude compensation is performed to a receiving signal. Then through a LS or MMSE algorithm, the pilot frequency channel information can be acquired and the channel estimations of all subchannels can be interpolated. Weighted average is performed to the channel estimations and the history channel information so as to acquire the response of the current channel. Then the response is applied for equalizing channel fading. An acquired symbol through decision is gone through amplitude limiting filtering so as to estimate nonlinear noise information. And the noise information is used for reestimation of the channel and signal detection. The method is suitable for a time-varying channel environment. Error diffusion can be avoided. Simultaneously, a disadvantage of the slow convergence speed by using the original method can be overcome.

The invention discloses a mobile robot path planning method based on an improved genetic algorithm. The mobile robot path planning method comprises: step one, carrying out environment modeling; to be specific, preprocessing a map by using a grid method and dividing an environment map into regular and uniform grids; step two, setting a start point and an end point of a robot; to be specific, setting the locations of the start point and the end point into free grids; step three, planning an optimal path of the robot based on an improved genetic algorithm; and step four, smoothening the path, determining a path point in the optimal path generated based on the algorithm as a control point, and carrying out smoothing on the optimal path by using a cubic B-spline curve to generate a final path. Therefore, technical problems of low efficiency, slow convergence speed, and great easiness of trapping in a local optimal value because of lots of iteration for optimal path calculation for the algorithm according to the mobile robot path planning method in the prior art are solved.

Based on principle of minimization risk of configuration, combined with cooperative collaboration evolution algorithm, and learning network structure of neural network and connection weight value, the invention obtains optimal compromise between network structure and connection weight value finally. The method includes three basic steps: data processing, network learning and network estimated forecast. Configuring network and learning connection weight value are carried out at same time in the invention so as to better solve practical problems existed in traditional neural network learning: correlation between result and initial value, slow convergence rate, easy to run to local minimum value as well as derivable error function needed and over learning. The invention raises learning capability and generalization capability of network, applicable to intelligent diagnosing heart disease, fault diagnosis in industries, stock and goods price forecasting etc.

The invention relates to an unmanned-ship speed and uncertainty estimation system and a design method. According to the system, an echo state network can be applied to speed estimation of an unmannedship, the echo state network is utilized to approximate model uncertainty and environment disturbance to enable the system to obtain target speed observation values, and also approximate unknown dynamics generated by the uncertainty of model parameters, non-modeling of fluid dynamics, external interference caused by wind waves and ocean currents and the like, and the state observation problem containing the model uncertainty and the unknown environment disturbance is effectively solved. Introduction of the echo state network overcomes the problems of slow convergence, proneness of falling intolocal minimums, complicated training processes and the like brought by traditional neural networks based on a learning algorithm of gradient descending. According to the system, the neural network with a low-frequency learning link is adopted to approximate system uncertainty, high-frequency oscillation which may be caused by a high-gain learning rate is effectively filtered out, and steady stateobservation on a system with unknown dynamics is realized.

The invention provides a feature matching method based on local structure similarity. The feature matching method is used for solving the problem that a matching result is not ideal due to noise interference in the image registration process. The method comprises the steps of 1, performing feature extraction and initial matching on two to-be-matched images; Step 2, establishing a neighborhood affine coefficient matrix of the feature points; Step 3, for each matching in the initial matching set, calculating the difference of the neighborhood affine coefficient matrixes of the feature points associated with the matching; 4, optimizing the neighborhood affine coefficient matrix to obtain a local structure difference degree; and 5, setting a comparison threshold value according to the local structure difference value of each matched and associated feature point, and determining a final feature matching pair as a matching relation result of the to-be-matched image. According to the method,the problems of complex optimization process and slow convergence in the prior art are technically solved, and the matching efficiency is effectively improved.

The invention, which belongs to the field of the information technology, relates to an energy consumption prediction method for improving an energy utilization rate of an iron and steel enterprise. A neural network algorithm with error feedback correction learning is used for carrying out prediction. In order to solve a problem of slow convergence speed of the traditional learning method, a momentum is added into an adjusting formula of a weight value and a threshold value, thereby accelerating a training speed of a network. With an adaptive adjusting learning rate, a learning rate of training at an error curved surface flat area is increased; and a learning rate is reduced at a drastic error changing area. On the basis of time-varying-weight-based parameter combination optimization method, distribution of an initial weight value and an initial threshold value is optimized, and good searching space is localized in solution space to avoid a local minimal point. According to the energy consumption prediction method, characteristics of energy consumption of all procedures are taken into consideration; the energy utilization rate is improved; and the energy costs are reduced and thus total production cost is lowered.

The invention discloses a full waveform inversion method based on preconditioned conjugate gradient acceleration algorithm, and belongs to the technical field of seismic exploration inversion imaging.According to the method, the shortcomings of slow convergence speed and low deep inversion accuracy of conjugate gradient full waveform inversion methods are overcome. The advantages of a gradient preconditioning operator and a Nesterov acceleration gradient descent method are combined, while the deep inversion accuracy of the conjugate gradient full waveform inversion method is improved, the convergence speed is effectively increased, thus the calculation amount of full-waveform inversion is reduced, and the technical support is provided for practical application of the full-waveform inversion.

The invention provides a water quality evaluation prediction method based on a fuzzy wavelet neural network and aims to solve problems of slow convergence speed during water quality prediction, poor approximation effect and inaccurate prediction result existing in a BP neural network in the prior art. According to the method, a fuzzy wavelet neural network prediction model is constructed through utilizing the known water quality analysis index quantity, prediction index quantity and fuzzy rule quantity, and the fuzzy wavelet neural network prediction model comprises an input layer, a subordinate layer, a fuzzy rule layer, a wavelet layer, an output layer and a defuzzication layer; a subordinate function parameter and a wavelet parameter of the wavelet layer are adjusted, a cost function is further defined, a BP algorithm based on a gradient descent method is utilized to carry out parameter adjustment, problems of low convergence speed, easy-to-generate concussion effects and local optimum can be avoided, model stability is improved, an initial parameter is optimized through employing an artificial bee colony algorithm, and the method is mainly applicable to water quality index prediction.

The invention discloses an improved ant colony optimization-based collision avoidance planning method for a USV in an unknown static obstacle environment, and belongs to the technical field of water surface unmanned ship collision avoidance planning. The method comprises the following steps: firstly, constructing a global coordinate system and a local coordinate system, and establishing a navigation radar simulation model; Designing a rolling optimization window for real-time collision avoidance planning; Then constructing an environment model by adopting a view method; Designing an improved ant colony optimization method for real-time collision avoidance planning; And finally, inputting navigation radar detection information and target point information into a USV static collision avoidance planner based on an improved ant colony optimization method, and obtaining an adjustment instruction of bow turning and speed of the USV at the next moment. According to the method, a rolling optimization window method and an improved ant colony optimization method are combined, so that the real-time performance of USV online planning is improved; For the slow convergence speed of the ant colony optimization method, an improved pseudo-random proportion rule is provided to select ant state transition; The global pheromone is updated by referring to the wolf pack distribution principle and the maximum and minimum ant system, so that the search is prevented from falling into local optimum.

The present invention provides a method for initialization and stepsize control of a time-domain equalizer (TEQ) in a receiver of a multi-carrier communication system to upgrade the performance of adaptive TEQ algorithms. As to TEQ initialization, the Time-domain Window Mask method generates a modified channel impulse response (CIR) by performing a locate maximum energy algorithm and then applies a time-domain window mask to adjust the modified CIR to obtain an initial value of a target impulse response. Then, a dividing operation is performed on the frequency-domain initial target impulse response and the modified CIR to determine an initial value of the frequency-domain TEQ impulse response. The Head-tail Equalizing method also performs the locate maximum energy algorithm, and the remaining points other than the consecutive points with maximum energy are combined and padded zero to the last few points to generate a modified CIR. The reciprocal of the frequency-domain modified CIR is determined to be an initial value of the frequency-domain TEQ impulse response. The initial value of the frequency-domain target impulse response is determined by multiplying the determined frequency-domain. TEQ impulse response with the frequency-domain of the CIR. The stepsize control method configures the stepsize coefficient as a time-varying coefficient during the whole adapting TEQ algorithm. It is smaller value at an early stage to prevent from divergence and becomes larger at a late adapting stage to prevent from slow convergence.

The invention relates to an orientation ant colony route optimization method based on evolution graph full route forecasting. The orientation ant colony route optimization method is mainly characterized in that on an application layer, a control center or a GPS provides a node track, link dispatching information or node positions and motion speed and direction information and sends the node track, the link dispatching information or the node positions and the motion speed and direction information to a network layer; on the network layer, an evolution graph model is firstly built by nodes according to information of the application layer, then full route forecasting is carried out, finally the orientation ant colony route optimization method is used for selecting an optimal route, and the data are sent; on a physical layer, link delay, data transmission rate and available bandwidth information is collected. The orientation ant colony route optimization method is reasonable in design, orientation ant colonies are used, slow convergence caused by frequent network topology changes is avoided, a routing algorithm meeting the QoS requirement is provided, and performance indexes are obviously improved compared with a traditional mobile Ad Hoc network DSR and AODV routing algorithm.

The invention discloses a voltage sag random prediction method based on actual power network monitoring information, belonging to the technical field of power quality analysis. The method comprises the following steps of: reading actual power network parameters to obtain node admittance matrix; processing the measured sag data of each node and obtaining the probabilistic model of fault type and fault duration; establishing a system fault information model; generating fault information by Latin hypercube sampling, and forming the original database of fault information; carrying out fault simulation to calculate the error between the predicted result and the measured result. According to whether the error meets the requirement, the fault information model is corrected until the error meets the requirement or reaches the preset correction times, and the final predicted result is obtained. The present invention can effectively avoid the problems of poor stability, slow convergence, long time consuming, large error of prediction result and the like in the existing prediction method, and the prediction result is more accurate.

The present invention provides a kind of power system environmental economic dispatching strategy based on improved multi-objective gravitational search algorithm, and the method comprises the following steps: 1. Constructing while considering that the minimum operating cost of the system and the minimum discharge of pollutants are objective functions, and establishing electric power System environmental economic optimization scheduling model; II. Established a power system environmental economic scheduling model that comprehensively considered system operating costs and pollutant discharge costs, and proposed an improved multi-objective gravity search algorithm to solve the model; III. For basic In order to solve the problem of slow convergence speed of the gravity search algorithm, in the process of updating the individual position, inspired by the particle swarm optimization algorithm, the position update formula of the gravity search algorithm has been improved; IV. In order to guide the group to approach the Pareto optimal solution set area and ensure the algorithm The solution set is evenly distributed, and the elite retention strategy is adopted; V. The fuzzy set theory is used to generate the best compromise solution, and a scheduling plan is provided for decision makers.

The invention provides a USV collision avoidance planning method based on improved ant colony optimization under dynamic obstacle online perception, and belongs to the technical field of collision avoidance planning of unmanned surface vehicles. For the problem of the insufficient search ability of a collision avoidance planning algorithm of a USV in a dynamic unknown environment, a real-time USVcollision avoidance planning method based on an improved ant colony optimization algorithm. Firstly, in order to meet <International Regulations for Preventing Collisions at Sea>, a reverse eccentricexpansion method is designed to expand dynamic obstacles; secondly, based on a motion velocity model and the improved ant colony optimization algorithm, USV collision avoidance planning in a known dynamic environment is converted into a multi-condition objective optimization problem in a known instantaneous static environment; lastly, for the problem of slow convergence speed of the ant colony optimization algorithm, an improved pseudo-random proportion rule was proposed to select ant state transition, and the global pheromone is updated by using a wolf pack allocation principle and a maximum-minimum ant system for reference, thereby avoiding the situation that the search falls into in local optimum.

The present application relates to tracking a position of a device, e.g. for detecting slow and rapid earth deformation, by making use of a recursive filter having the filter characteristic adapted to a detected type of motion. If the motion of the position tracking device is rapid, the filter characteristic is set such that the rapid motion can be tracked with the necessary speed. On the other hand, if the motion is slow, e.g. during times of a normal tectonic drift, the filter characteristic is set such that the motion is slowly tracked with the advantage of efficient noise reduction, i.e. noise in the input signal is effectively barred and does not pass through the filter to the output signal. Thus, in times of rapid motion the convergence speed of the filter output signal to the input signal is set high for fast convergence and in times of slow motion the convergence speed of the filter output signal to the input signal is set low for a slow convergence. The filter may be a Kalman filter.

The invention relates to an intelligent optimization method for ship dynamic positioning thrust distribution. The method comprises steps that S1, the layout of a ship propulsion unit is determined, parameters of each thruster are loaded, and the parameters of the thruster include the variable range of the thruster thrust direction, the variable range of the thrust magnitude, the change rate rangeof the thrust direction and the change rate range of the thrust magnitude; S2, change rate restriction of a thruster state vector is acquired according to the parameters of each thruster and a presentstate, after a target control instruction is received, a multi-step optimization problem model is established; S3, a global optimal solution of a state of the thruster in the long-term change range is acquired through the genetic algorithm; and S4, an optimal multi-step decision sequence for driving the thruster is acquired, and distribution decisions are outputted. Compared with the prior art, problems of slow convergence and rapid degradation of population diversity of the genetic algorithm are solved.

The invention discloses an electric energy quality disturbance positioning and identifying method based on the lifting of the wavelet and the improvement of a BP neural network. The method comprises the following steps: using the Euclidean decomposition algorithm to obtain a db4 wavelet lifting scheme; performing lifting wavelet decomposition to the disturbance signal; in combination with the modulus maxima, performing the positioning detection to the disturbance mutation point peak; and utilizing the method of combining the self-adaptive learning rate with the incremental momentum term to improve and carry out disturbance identification to the BP neural network. The method of the invention can better obtain the disturbance period information, achieves fast positioning and high precision, and overcomes the shortcoming that a traditional BP neural network is prone to fall into local minimal points and slow convergence speed, therefore, achieving a high identifying rate for the power distribution network system's electric energy quality disturbance.

Disclosed in the invention is a support-vector-machine-regression-based method for predicting a wind speed of a wind power plant. The invention relates to a support-vector-machine-regression-based method for predicting a wind speed of a wind power plant, thereby solving problems of low prediction precision and slow convergence speed of the existing wind speed prediction method. The method comprises the following steps: step one, selecting sample data collected by a wind power plant; step two, determining a sample training set and a testing set; step three, carrying out pretreatment on the sample data; step four, selecting a support vector machine (SVM) regression kernel function and determining a to-be-optimized parameter of an SVM model; and step five, training the SVM model by using an optimal parameter and predicting a wind speed value at a future time. The method is applied to the wind power prediction field.