70results about How to "Solve estimation problems" patented technology

The invention discloses an automatic driving overtaking decision-making method based on reinforcement learning under opposite double lanes. The method comprises the following steps of collecting the traffic state of an automatic driving vehicle through a sensor; inputting the collected traffic state into a trained decision model; enabling the decision model to select a corresponding driving actioninstruction from the action space according to the input information and output the driving action instruction, and driving the vehicle automatically to form a new traffic state after the driving action; calculating a reward value of the driving action through a reward function, and storing the original traffic state, the driving action, the reward value and the new traffic state into an experience playback pool as transfer samples; calculating a loss function value of the decision model, and optimizing parameters of the decision model according to the transfer sample and the loss function value; and repeating the steps until the automatic driving is finished. The safety and comfort of the overtaking decision-making process of the automatic driving vehicle are ensured, and the humanization and robustness of decision-making are improved through a reinforcement learning decision-making method.

The invention discloses a multi-Bernoulli video multi-target detection tracking method based on YOLOv3, and belongs to the field of machine vision and intelligent information processing. According tothe method, a YOLOv3 detection technology is introduced under a multi-Bernoulli filtering framework, an anti-interference convolution feature is adopted to describe a target, a detection result and atracking result are interactively fused, and accurate estimation on the multi-target state of the video with unknown number and time varying is realized; in the tracking process, a matched detection frame is combined with a target track and a target template, target newborn judgment and occlusion target re-identification are carried out in real time, identity label information of a detection target and an estimation target is considered at the same time, target identity identification and track tracking are achieved, the tracking precision of an occluded target can be effectively improved, andtrack fragments are reduced. Experiments show that the method has a good tracking effect and robustness, and can widely meet the actual design requirements of systems such as intelligent video monitoring, man-machine interaction and intelligent traffic control.

The invention provides a lithium iron phosphate battery modeling and SOC estimating method in account of capacity loss. The method is provided based on the facts that the working state of the existing lithium iron phosphate battery is influenced by many factors of temperature, current, cycle times, depth of discharge and the like, and thus the modeling process is very complicated. According to the method, based on the Thevenin equivalent circuit, the modeling work is conducted on the lithium iron phosphate battery; the model open-circuit voltage, resistance and capacitance value identification method are given; and the capacity loss of the lithium iron phosphate battery in the life cycle is taken into account; a capacity estimation mathematical model is established; the output precision of a lithium iron phosphate battery model is improved; and the lithium iron phosphate battery SOC estimating problem brought by uncertainty noise is solved with extended Kalman filtering (EKF) algorithm. The method has the advantages of being simple, scientific and reasonable, high in application value, good in effect and the like.

The invention discloses a battery state-of-charge (SOC) estimation method based on nonlinear prediction extended Kalman filtering (NPEKF). The method has the advantages of being simple in operation and high in precision. The method includes the steps that (1) time needed for the whole method is averagely divided into N time periods, each time period stands for one step, in another word, the kth time period stands for the kth step, k is less than or equal to N, and both k and N are positive integers; (2) a model of a battery system is established according to the Kirchhoff voltage and current theorem, and an error matrix d (k) of the model of the battery system and a model error allocation matrix G(k) of the battery system are obtained; (3) compensation is carried out on a prior state estimating equation obtained in the step (2); (4) a posterior state estimation equation of the battery system in the (k+1)th step is solved to obtain an SOC value; (5) according to the posterior state estimation result, obtained in the step (4), of the battery system in the (k+1)th step, the SOC value is compared with a true SOC value of a battery, effectiveness of an NPEKF algorithm is verified, k=n+1, and the step (3) is repeated until the Nth step is executed.

The invention discloses a sparse construction and reconstruction method of array space signals. Due to the fact that the angle of received incident signals changes constantly under the high dynamic condition, enough sampling signals can not be acquired to conduct traditional beam forming. According to the method, a gain test of the array space signals is conducted through the Capon beam forming method based on feature space decomposition, and establishment of a small-snapshot-number sparseness model and OPM reconstruction of the small-snapshot-number sparseness model are conducted based on the compressed sensing theory. The result shows that the method has the advantage of processing undersampled data and achieves estimation on the angle of arrival of the array signals. The method further shows that in the space signal processing field, sparsity makes extraction of related information faster and more effective, so that the signal acquisition cost and the signal processing cost are lowered. By means of the method, expected waveform gain can be obtained in the expected direction of arrival. Compared with a direct sampling method, the method can solve the small snapshot number problem occurring under the high dynamic condition and improve the robustness of a system.

The invention discloses a wireless communication link estimation method facing a high-speed railway linear community, and belongs to the technical field of wireless mobile communication. The method comprises the following steps: 1) a set of classification standard aiming at a high-speed railway characteristic spreading environment is put forward; 2) modeling is performed on large-scale path loss and shadow fading under various scenes, and a set of standardized spreading model library facing a high-speed railway is established; and 3) a coverage prediction method facing the railway linear community is put forward. The coverage range of the high-speed railway community is more accurately predicted under the premise of being without preliminary measurement so that a network arrangement scheme of the high-speed railway is optimized, base station spacing distance among communities is reasonably arranged and thus investment in high-speed railway wireless communication infrastructure is greatly reduced.

The invention discloses a traffic flow estimation method based on the network charging data of a highway. The method is conducted based on the historical network charging data of the highway. The method comprises the steps of firstly, estimating to obtain the traffic flow of each road section of a road network within each of different time periods in history; analyzing to obtain the distribution mode of the traffic flow at each inlet of the road network within a plurality of time periods in the past; based on the position and the time for each of all current in-transit vehicles to enter the road network, completing the estimation on the real-time traffic flow of the highway road network. According to the method for extracting the distribution mode of the traffic flow, the contribution rate of one O relative to the D within a time period t is recorded as COD, and the stability mode of the COD is measured based on the information entropy so as to obtain the distribution mode of the traffic flow. The method is low in cost and can cover the entire road network. Meanwhile, the method has the advantages of all-weather operation, high data quality and the like. The estimation problem of the traffic flow for the highway road network is solved without any additional hardware cost. Therefore, the method serves as an effective means and a beneficial supplement for monitoring the traffic flow of the road network in the prior art.

The invention relates to a high-voltage single conductor cable service life evaluating method and system. A high-voltage single conductor cable service life evaluating method comprises the following steps of acquiring high-voltage single conductor cable original operation zero-failure data, conducting estimation for shape and size parameters of a double-parameter Weibull distribution model built for evaluating high-voltage single conductor cable service life via a corrected maximum likelihood function to acquire shape parameter estimation values and size parameter estimation values, and acquiring a high-voltage single conductor cable service life distribution probability curve via the double-parameter Weibull distribution model according to the shape parameter estimation values and size parameter estimation values. Accurate fault analysis and fault prediction can be provided with fewer samples by the use of the Weibull distribution model; parameter estimation can be achieved even when zero-failure samples do not exist by the corrected maximum likelihood function; and accuracy for insulation cable service life estimation can be improved.

The invention provides a polarized bistatic MIMO radar parameter estimation method in an impact noise environment based on sparse reconstruction. The method comprises the steps: building a polarized bistatic MIMO radar receiving data model in the impulsive noise environment; 2, carrying out the impact removing preprocessing of snapshot sampling data; 3, carrying out the estimation of transmitting angle parameters of a polarized bistatic MIMO radar through the rotational invariance of a subspace; 4, constructing a sparse dictionary set of the polarized bistatic MIMO radar; 5, carrying out the sparse reconstruction of a receiving angle of the polarized bistatic MIMO radar; 6, judging whether to reach the maximum number of iteration times or not: executing step 7 if the maximum number of iteration times arrives, or else supposing k=k+1, and returning to step 5; 7, obtaining a sparse reconstruction result sk, and obtaining the information of the receiving angle of the polarized bistatic MIMO radar through an index set U, and outputting an estimation result of the transmitting angle and the receiving angle of the polarized bistatic MIMO radar. The method is wider in application range, and can be used for solving actual problems that cannot be solved by a conventional polarized bistatic MIMO radar parameter estimation method.

The invention provides a malicious web crawler detection method based on a hidden Markov model (HMM), and belongs to the technical field of computer software. A malicious web crawler maliciously captures sensitive information and privacy data of a website without permission, and meanwhile a barbaric traffic behavior mode has a negative effect on the service quality of the website. An existing web crawler detection method cannot accurately recognize the malicious web crawler, and has a relatively high misjudgment rate. The novel malicious web crawler detection method based on the HMM provided by the invention specifically comprises the following steps: (1) performing modeling based on the user HTTP (Hyper Text Transport Protocol) traffic of the HMM; and (2) performing modeling based on web crawler behaviors of an HTTP.

The invention provides a wireless communication method, a terminal device and a network device. Candidate demodulation pilots of control channels in different search intervals have quasi-co-location QCL association with different types of pilot signals, and the problem of channel estimation during the blind detection of a control channel is solved. The method comprises the following steps: the terminal device determines a pilot signal corresponding to at least one search interval in a plurality of search intervals of the control channel, wherein the pilot signal corresponding to each search interval in the at least one search interval has quasi-co-location QCL association with the candidate demodulation pilots of the control channel in each search interval, and the types of the pilot signals corresponding to different search intervals in the plurality of search intervals are different; and according to the QCL association between the pilot signal corresponding to each search interval and the candidate demodulation pilots of the control channel in each search interval, the terminal device detects the candidate control channels in the at least one search interval.

The invention discloses a traffic matrix estimation method which is based on combination of a linear programming method and Zernike-Moment. According to the traffic matrix estimation method, due to adoption of the linear programming method, an objective function is selected instead of prior information, and therefore the sensibility of a model on the prior information is eliminated. According to interpolation processing of the Zernike-Moment method, the problem of massive data loss which cannot be solved in a K-NN (K-Nearest Neighbor) algorithm is solved, the sensibility of the model on a missing value is eliminated, and meanwhile, the influence of noise in input information on an estimation result is eliminated. In addition, due to adoption of the Zernike-Moment method, a result of the linear programming method can be corrected by utilizing a traffic matrix estimation value obtained by using the LP (Linear Programming) method as well as structural similarity, redundant information and space-time restriction information among traffic matrix data, and furthermore the estimation accuracy of the traffic matrix is improved.

The invention relates to the field of channel models and channel parameter estimation algorithms, and provides a new scattering cluster model channel parameter estimation method having the characteristics of being high in accuracy, decreased in calculated quantity and reduced in complexity. The technical scheme provided by the invention is as follows: the improved MUSIC algorithm scattering cluster model channel parameter estimation method has the following steps of 1, building a system model, wherein a frequency array response array B(f) is L frequency sampling in short time, fs indicates a sampling rate, FORMULA (the formula is as shown in the specification), wherein S(T-[tau])=[S(t-[tau]1), ..., S(t-[tau]D) ]<T> indicates a transmitting signal with delay [tau]; and 2, improving a MUSICalgorithm in the system model, including 2.1, using a time domain filtering method, and 2.2, using an airspace smoothing method to acquire a frequency spectrum: FORMULA (the formula is as shown in thespecification). The method provided by the invention is mainly applied to the channel models and channel parameter estimation occasions.

The invention relates to a distribution multi-input multi-output (MIMO) system and relative channel evaluate method, wherein said system comprises sender and receiver; the sender comprises distribution MIMO code module, and multiply parallel sending circuits for inserting pilot sequence and adding protective distance into dataflow output from coding module, to form and send data frame; the receiver comprises multiply parallel receiving circuits for receiving the data frame, obtaining the pilot sequence and data section from the data frame; the pilot mark module for outputting the number of sending circuit and the pilot sequence at preset length; channel evaluating module for obtaining channel evaluating matrix via the pilot sequence of receiving circuit and the pilot sequence of pilot mark module. The invention can process channel evaluation of distribution MIMO system.

The invention discloses a coprime array DOA estimation method based on a quantum penguin search mechanism under strong impact noise. The method comprises the following steps: obtaining a fractional low-order covariance matrix of an infinite norm weighted signal; constructing an extended fractional low-order covariance matrix and an extended guide matrix, and constructing a maximum likelihood direction finding equation based on the extended covariance matrix; initializing a quantum penguin group, and initializing the speed and the quantum position of the quantum penguin; calculating fitness andcarrying out descending sort, and selecting a global optimal quantum position; updating the quantum position of the quantum penguin; evaluating a quantum penguin position fitness value, and updatinga global optimal quantum position; judging whether the maximum number of iterations is reached or not, and if not, continuing iteration; and if so, outputting a global optimal quantum position, and mapping to obtain an optimal DOA estimation value. According to the invention, higher direction finding precision, higher operation speed, better array aperture expansion capability, coherent source measurement capability, impact noise resistance and wider application environment are realized.

The invention, which belongs to the field of array signal processing parameter estimation, particularly relates to a uniform-circular-array-based coherent signal parameter estimation method in an impact noise environment. The method comprises the following steps that: step one, snapshot sampling is carried out on D source signals in space; step two, de-impact preprocessing is carried out on the snapshot sampling data; step three, mode excitation transformation is carried out on array output data; step four, a sparsely reconstructed dictionary set is constructed; step five, sparse reconstruction is carried out to obtain a coherent signal source azimuth; step six, whether a maximum iteration number is reached is determined; if so, a step seven is performed; and if not, t is enabled to be equal to the sum of t and 1 and the step five is preformed again; and step seven, a sparse reconstruction result is obtained, source azimuth information is obtained by an index set U, and a coherent source direction-of-arrival estimation result is outputted. Therefore, a uniform-circular-array-based coherent signal parameter estimation problem in an impact noise environment is solved and the mode excitation transformation and compressed sensing sparse reconstruction ideas are used as parameter estimation bases. The designed method has advantages of low computational complexity, short calculationtime and high robustness.

The invention relates to an energy perception based method for a cognitive user to estimate parameters of an authorized user in a cognitive radio communication system, belonging to the technical field of parameter estimation for authorized users. The invention aims at solving the problems that in the cognitive radio communication system, the cognitive user cannot obtain parameters of the authorized user during spectrum handover in appearance of the authorized user, and further cannot maximize the benefit. The method comprises that an authorized channel model is established, and the state probability distributions of authorized channels are obtained; busy and idle state changes of the authorized channels are described through a hidden Markov model (HMM), and a hidden state transfer probability matrix of the authorized channels is obtained; the detection probability and the false alarm probability are obtained via calculation in an energy perception method; a confusion matrix is further obtained; and finally the HMM is estimated in a Baum-Welsh algorithm to obtain the estimated value of the HMM. The method of the invention is used for estimating the parameters of the authorized users.

The invention discloses a mechanical vibration source number estimating method in an underdetermined blind separation; the method includes steps of applying a second-order non-stable source separation, separating the centralized sensor data to nonoverlapping data blocks; calculating a single time delay covariance matrix of every data block and overlapping to be three-order tensor or a parallel factor model; then applying a nuclear consistent diagnosis algorithm and estimating the best component number of the parallel factor model, and thereby acquiring the vibration source number of a mechanical system. Compared with the existed mechanical vibration source number estimating method, the presented method can effectively solve the vibration source number estimation problem under the underdetermined blind separation, and has wide application prospect in multi-fault composite diagnosis.

The invention discloses a self-adaption LOS guide method of hovership path tracking based on sideslip compensation, and belongs to the field of hovership control. The method comprises the following steps of 1, building a hovership three-degree-of-freedom motion mathematical model; 2, building an error dynamic model; 3, designing a self-adaption sideslip angle estimator; 4, designing a self-adaptation LOS guidance law of sideslip angle compensation; 5, designing a PID heading controller; 6, conducting simulation verification. By means of the self-adaption LOS guide method of hovership path tracking based on the sideslip compensation, the self-adaption estimation of the sideslip angle is applied to LOS guide of a hovership, the LOS guidance law is subjected to sideslip compensation, the self-adaption method is adopted to estimate the sideslip angle online, the estimation problem of the sideslip angle is solved, the problem that the path tracking effect is influenced when a hovership sideslips while sailing on the water surface and interfered by the external environment is solved, and the precision of hovership path tracking is improved.

The invention provides a channel estimation method of a multi-path non-Gaussian noise channel, and designs an empirical likelihood estimation method combining training sequence data and observation data. The method comprises the following steps of: firstly obtaining the observation data passing through the non-Gaussian noise channel from a signal receiving end, and generating an auxiliary variable in combination with the training sequence data; secondly, converting the estimation problem into a non-parametric problem according to the empirical likelihood theory proposed by Owen, namely that a non-parametric empirical likelihood variable under certain limiting conditions is generated by use of the auxiliary variable, and the likelihood variable is obtained by the Lagrangian method; and finally, obtaining empirical likelihood values corresponding to different channel estimation values by the Newton iteration method, and taking the channel estimation value corresponding to the maximum empirical likelihood value. Take the mixture of additive white Gaussian noise and pulse noise as an example, the MSE (mean square error) and BER (bit error rate) effects of multi-path channel estimation are very good.

The invention discloses a compressed sensing based two-layer heterogeneous network interference channel iterative estimation method, and belongs to the technical field of mobile communication. The compressed sensing based two-layer heterogeneous network interference channel iterative estimation method is characterized by comprising the steps of carrying out compressed sensing sampling for a received signal and removing a cyclic prefix by utilizing sparsity of a broadband wireless channel; carrying out initial estimation for expected data by adopting MMSE, regarding a sparse binary matrix of a deterministic structure as a measurement matrix; carrying out compressed sensing channel estimation by combining pilot symbols with same amplitude and uniform phase distribution; and removing the interference portion obtained through estimation from the signal and updating the expected data for cyclic iteration. By combining a compressed sensing framework with a channel estimation technology of a two-layer heterogeneous network, the interference channel information and the expected data information can be estimated at the same time after a small number of iterations; and compared with a traditional channel estimation method, the compressed sensing based two-layer heterogeneous network interference channel iterative estimation method has obvious performance advantages in the presence of high signal-to-noise ratio.

The invention discloses a dressing human body shape estimation method based on a deep neural network, and the method comprises the steps: constructing a three-dimensional clothes grid model database with abundant clothes types, and training a deep neural network model for predicting the fitness of a human body grid based on the database. According to the method, the problem of estimating the realbody shape of the user under complex dressing is solved; clothes types, grid geometric structures and human body posture actions are comprehensively considered, accurate detection results are providedfor complex human body actions and clothes types, the method is not limited to input of a grid sequence any more, a three-dimensional model of a single human body can also be input, and application scenes of body shape estimation are expanded.

The invention discloses a method for processing signals based on frame structure extension. The method comprises the steps as follows: a simulated transmission terminal sends uplink signals to base stations according to the universal frame structure or optimized frame structure of a network emulsion system; and the base stations process the signals according to the universal frame structures or optimized frame structures of the uplink signals. The invention also discloses a system for processing signals based on frame structure extension. By using the method and the system, the problem of channel estimation in case of long air interface propagation delay is solved; and the signals can be prevented from interfering with other time slots after being transmitted by the simulated transmission terminal, and can keep consistent with the standard frame structures to the greatest extent at the same time.

The invention discloses a monostatic MIMO radar DOA estimation method under an impact noise background. The method comprises the steps: step one, carrying out impact noise modeling based on S alpha Ssteady-state distribution and impact noise suppression based on a Cauchy kernel function; step two, carrying out signal subspace construction based on an iterative convex optimization algorithm; and step three, realizing DOA estimation based on a rotation invariant method. According to the invention, the iterative solving of array receiving data signal subspace is realized by using a complex-valued Newton gradient descent algorithm to solve a solution problem of the Cache kernel function; and after multi-times iteration, the signal subspace with good performance is obtained. The suppression ofthe impact noise is realized by using the Cache kernel function; and a DOA estimation problem under the background of the impact noise in the monostatic MIMO radar is solved. The azimuth parameter information estimation result is relatively accurate; and good incoming wave azimuth estimation is realized while spectral peak search is avoided.

The invention discloses a state space decomposition-based linear constraint estimation method. The method comprises the following steps of 1) setting a random linear discrete time system; 2) performing state space decomposition on each time point, and rebuilding a system model with a linear constraint state according to the set random linear discrete time system in the step 1) and a state space decomposition result; 3) calculating a propagation function of an unconstrained Sigma point; and 4) for a constrained system, limiting the Sigma point in a constraint domain defined in the specification, namely, when the Sigma point does not obey the rebuilt system model with the linear constraint state in the step 2), projecting the Sigma point to a boundary of the constraint domain defined in the specification, obtaining a Sigma point under interval constraints according to the calculated propagation function of the unconstrained Sigma point in the step 3), setting a weight of the Sigma point under the interval constraints, obtaining a propagation function of a constrained unscented point, and performing state space decomposition-based linear constraint estimation. The method can realize the linear constraint estimation.

The invention provides an authorization user parameter estimation method based on cooperation perception in a cognitive radio system and relates to the parameter estimation field in cognitive radio communication systems. Thus, cognitive users are capable of estimating authorization user parameters in a condition of low signal-to-noise ration in a cognitive radio system. Based on cooperative detection, the hidden Markov model is used for describing the changing process of busy and idle states of authorization users in an authorization system, so the matrix of the hidden state transition probability in the hidden Markov model is expressed by the arrival rate and the service rate of the authorization users in the authorization system. Each of the cognitive users transmits detection results to a cognitive base station which will process the collected data via the Or-principle so as to get the detection results. At last, the Baum-Welsh algorithm is used for estimating the hidden Markov model and getting the hidden transition probability and confusion matrix, thereby estimating parameters of the arrival rate and the service rate in the authorization system. The authorization user parameter is applicable in cognitive radio systems.

The present invention provides a uniform circular array-based DOA estimation and mutual coupling correction method. According to analysis and derivation, an actual steering vector of a uniform circular array in a mode space domain can be equivalently regarded as a steering vector of a virtual uniform linear array with channel amplitude and phase disturbance, which means that the problem of correcting the mutual coupling of the uniform circular array can be transformed into the correction of the channel amplitude and phase disturbance of the virtual uniform linear array: the amplitude disturbance is obtained according to the main diagonal of the covariance matrix of the data received by the virtual uniform linear array, and the phase disturbance is obtained according to the characteristic that the covariance matrix of the data received by the uniform linear array without amplitude and phase disturbance has a Toeplitz structure. According to the uniform circular array-based DOA estimation and mutual coupling correction method provided by the present invention, the problem of fuzzy estimation in phase disturbance estimation is solved, online work can be supported, the method has an analytic solution and no multi-dimensional parameter search, and the problem of false DOA estimation that appears in the RARE theory can be avoided.

The invention belongs to the technical field of reliability statistics, and discloses a Wiener process parameter estimation method based on performance degradation data and life data fusion. First, the initial value of Wiener process parameters is obtained by maximum likelihood method based on performance degradation data. Then, the failure probability of life data is estimated by using the initial value. Finally, the life distribution parameters of life data and failure probability are fitted and transformed into the final value of Wiener process parameters. The invention solves the parameterestimation problem of the Wiener process based on the fusion of the performance degradation data and the life data through the steps. Compared with the existing method, the algorithm calculation process proposed by the invention is simple and effective. As will be seen from the follow detailed embodiment, the algorithm proposed by the invention has high computational accuracy.

The embodiment of the invention discloses a signal-to-noise ratio estimation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a spreadspectrum complex signal received by a receiver, wherein the spread spectrum complex signal is a constant-amplitude signal; performing despreading processing on the spread spectrum complex signal to obtain a despread signal corresponding to the spread spectrum complex signal; wherein the spread spectrum complex signal and the dispreading signal are sampling signals and respectively comprise a plurality of sample point signals; determining useful signal power matched with the despread signal according to the signal amplitude information of the sample point signal in the despread signal; and estimating the signal-to-noise ratio of the spread spectrum complex signal according to the total power of the spread spectrum complex signal and the useful signal power. The method can reduce the operation complexity of signal-to-noise ratio estimation.

The invention discloses a grey bootstrap and unascertained rational number-based small sample test data estimation method. The method comprises the following steps of S1, generating grey bootstrap ofsmall sample test data; S2, constructing and optimizing an unascertained rational number: establishing the unascertained rational number, and when a credibility entropy is a maximum value, estimatingan optimal order of the unascertained rational number, thereby realizing optimization, wherein the larger the credibility entropy of the established unascertained rational number is, the better the description of test indexes is; and S3, performing parameter estimation based on the unascertained rational number. According to the grey bootstrap and unascertained rational number-based small sample test data estimation method, point estimation, interval estimation and estimation reliability models are given, and example verification is performed; and a simulation example shows that the method isreasonable and feasible, the parameter estimation problem of equipment test data can be effectively solved, and a probability distribution feature of original data is not needed.