22 results about "Least squares" patented technology

The invention discloses a pilot frequency distribution method in DCS channel estimation. The method comprises modeling a pilot frequency position optimization problem into a combinational optimization problem with the minimum channel estimation error serving as the starting point; solving the combinational optimization problem through a raised genetic algorithm to obtain an optimal pilot frequency position set enabling the channel estimation error to be the smallest. By the aid of the method, compared with traditional least square channel estimation, lower estimation errors and higher frequency spectrum effectiveness can be obtained on the basis of DCS.

The invention discloses a baseband receiving method and a baseband device applied to a terminal in a narrow-band wireless system, and relates to wireless communication technology. The baseband receiving method includes the steps: calculating LS (least squares) estimation values of an NRS (numeric rating scale) pilot frequency position, averaging LS results of the same subcarrier position in a subframe, and encrypting LS results of different carrier positions to obtain a list of results; calculating frequency domain correlation values, and distinguishing different gears of the correlation values. The list of encrypted LS results are filtered according to the distinguished gears, filter results serve as channel estimation values of all OFDM (orthogonal frequency division multiplexing) symbols on the subframe, noise power values are estimated according to the encrypted LS results, and finally, received signals are subjected to MIMO (multiple input multiple output) detection and demodulation according to the channel estimation values and the noise power estimation values. Compared with the prior art, the baseband receiving method has the advantages that the calculating amount of the baseband receiving process can be greatly reduced, the method is suitable for the narrow-band wireless system with low signal-to-noise ratio and low movability, and good receiving performance is achieved.

The invention discloses a training sample grouping construction method used for support vector regression (SVR) short-term load forecasting, and belongs to the field of intelligent computing and machine study. The training sample grouping construction method comprises a step of analyzing correlation, wherein the correlation degree of the load of each time interval and the loads of other time intervals is analyzed through the Tangs correlation degree of the grey correlation degree to form a correlation degree matrix; a step of grouping prediction problems, wherein the time intervals with high load correlation degree are divided into one group according to the correlation degree matrix; a step of constructing a reference load matrix; a step of selecting a reference load to construct a training sample, wherein linear function fitting is carried out on each row of the loads in a load variation rate matrix in a least square fit mode, and fitting variance is calculated; and a step of selecting the load of the time interval with small fitting variance to serve as the forecasting reference load of the group. The training sample grouping construction method used for the SVR short-term load forecasting is capable of improving the load forecasting accuracy, and avoids the problem of high time complexity. The experiment result shows that a short-term load forecasting model trained by the training sample constructed through the method has good performance in forecasting accuracy and time complexity.

The invention discloses a method for improving a microalgae cultivation condition by using metabonomics so as to improve the oil-producing capacity. The method comprises the following steps of: (1) detecting intracellular metabolites in microalgae; (2) analyzing the oil-producing capacity of the microalgae; (3) performing partial least squares (PLS) analysis; and (4) process analysis. The oil-producing process of the microalgae, an intracellular metabolism change rule and important metabolites in the oil-producing process can be researched wholly by the method, and the change rule of the metabolism levels lays a base for the internal mechanism of the oil-producing process, so a theoretical base is laid for further optimizing a microalgae cultivation process and improving the grease yield. A new concept and a new method are provided for the research on other oil-producing microorganism cultivation processes.

The invention discloses a method for measuring river-crossing levelling altitude difference by using a GPS (global positioning system), which comprises the following steps: selecting at least two groups of control network points (two in each group) at two sides of a river in a GPS engineering control network as river-crossing levelling points; measuring the GPS geodetic elevation of each point at two sides of the GPS engineering control network to obtain the GPS geodetic altitude difference between the river-crossing levelling points; levelling to obtain the levelling elevation of each control network point of the known river, assuming that the levelling altitude difference between the two river-crossing control network points in one selected group is unknown, and calculating the known levelling elevation of the other river control network point containing the unknown; building a curved surface or plane fitting model according to the relationship between the GPS geodetic elevation and the levelling elevation at the same control network point; and calculating the levelling elevation and river-crossing levelling altitude difference of the assumed river-crossing levelling points by using a least square method. By using the method to measure the river-crossing levelling altitude difference, the workload is reduced, and the working efficiency is improved. Meanwhile, the calculated river-crossing levelling altitude difference has high stability, and the result is reliable.

The invention provides a neural network increment-type feedforward algorithm based on sample increment driving. According to characteristics of input samples, input weights and hidden layer thresholds of a neural network are obtained by a least square method, and initialization of model parameters is completed; and the input weights and hidden layer thresholds of the model are updated according to increments between new samples and old samples, function relations between the model parameters and input samples are established, and functions of sample self-adaptability and online feedforward adjustment are realized. The algorithm provided by the invention has the advantages that the prediction precision is high, the generalization capability is high, and online feedforward adjustment is available.

The invention relates to a method for processing mass concentration data of a suspended sediment based on a laser particle size analyzer. The method mainly comprises the following steps of acquiring volume concentrations of suspended particles with different particle sizes at each site of an observation sea area by use of the laser particle size analyzer; performing mass control on the original data of the acquired volume concentrations of the suspended particles with different particle sizes; performing linear fitting according to the mass concentration of the suspended sediments measured bya chemical weighing method at partial sites and the total volume concentration of the suspended particles at corresponding positions by use of a least square method to acquire a fitting coefficient and a correlation coefficient; and fitting the mass concentration of the sediment at each site according to the total volume concentration of a suspended matter at each site of the surveyed sea area byuse of a fitting relation acquired by linear fitting.

The invention discloses a channel estimation algorithm of an IQ imbalance OFDM system under a sparse environment. A pilot symbol passes through a channel under the IQ imbalance condition, and a receiving end signal is obtained. IQ imbalance is taken as a channel parameter to be brought in estimation, sparse characteristics are utilized, and a time domain channel impact response is estimated through a least square method. A proper threshold is determined by estimating noise variance, and noise within cyclic prefix length is filtered out. According to the threshold time domain least-square estimation method based on noise variance estimation, good compromise of performance and complexity is achieved, meanwhile, only one OFDM symbol is needed, the high spectrum utilization rate is achieved, and system error code performance can be effectively improved.

The invention discloses a method for accurately estimating a spatial channel system function based on a sparse theory, which comprises the following steps of: establishing a sparse regular least square model based on a mixed l0 norm when channel estimation is carried out by using received data in consideration of the fact that the number of main channels in a scene is limited; in consideration ofthe non-convexity of the problem, using a convex optimization theory to carry out convex relaxation on the non-convexity, and giving a sparse regular least square model of a mixed l1 norm; and then reconstructing the received data according to a solving result, and adjusting a weight factor between a fitting error and channel sparsity in the optimization model according to an error between the received data and the original data until the weight factor meets a convergence condition. According to the method, the priori information of spatial channel sparsity can be fully utilized, so that the receiving data fitting error and the sparsity of the channel coefficient vector are optimized, the method better conforms to the actual situation, and the deviation of channel coefficient estimation caused by only considering the fitting error when a traditional method is used for estimating the spatial channel is effectively avoided.

The invention belongs to the technical field of artificial intelligence, and discloses a dynamic width-depth neural network for hyperspectral classification and a learning method thereof. According to the network, an iterative least square method is used when an EWSWS layer carries out extended learning on advanced features in the width direction and the depth direction, a plurality of EWSWS networks are added into the network one by one, and more learning ability is obtained in an increment mode. When training data learning is sufficient, the growth process automatically stops. In the process, training data are segmented in a feature domain and a sample domain, so that the method can be used for hyperspectral image data with high feature dimensions and a large number of training samples. Another advantage of the network is that when a proper system structure is gradually and automatically searched, the learning process is more stable, and very specific hyper-parameters and trained weights can be given in a full connection layer.

The invention discloses an indoor positioning method based on four-dimensional code mapping, and the method comprises the steps: (S100), carrying out the Kronecker product of a horizontal-dimensionalfeature vector matrix and a vertical-dimensional feature vector matrix, generating a feature vector matrix of an antenna group, mapping the feature vector matrix into a four-dimensional code, carryingout the matching, obtaining an approximate code word four-dimensional code, and obtaining a pre-coding matrix; (S200), performing beam forming on the base station transmitting signal according to theprecoding matrix of the mobile terminal; and (S300), forming a virtual cell taking the mobile terminal by the to-be-positioned mobile terminal and the antenna group as the center; and forming a van virtual cell taking the mobile middle end as the center with surrounding mobile terminals, constructing an estimation function to obtain a positioning target function, solving the positioning target function by adopting a nonlinear least square algorithm, and taking the coordinate when the positioning target function takes the minimum value as the coordinate of the to-be-positioned mobile terminal.According to the method, beam matching of four-dimensional code mapping is combined with a virtual cell cooperation mode, and indoor real-time space positioning is achieved.

The invention discloses an emotion-based indoor space lighting system influence factor modeling method, which comprises the following steps of: selecting illumination, color temperature and color rendering indexes as influence factors of an indoor lighting system, performing single-factor control variable experiments, establishing a matrix of emotion feedback of a user to a lighting environment in each group of experiments, and establishing a relation curve between the influence factors and the emotion and a fitting equation between the influence factors, and finally using a least square method to fit a k-order curved surface equation of illumination, color temperature, color rendering index and emotion to determine an optimal solution and an optimal interval range. The color rendering index is introduced as an influence factor when the relationship between the lighting system and the user emotion is researched, so that the model is more scientific and objective; mathematical means are used for analysis and prediction in industrial design, the design reasonability of the lighting system is evaluated, and subjectivity and fuzziness of manual judgment are avoided.

The invention discloses a single-ended direct-current distribution line single-pole grounding fault distance measurement method, and belongs to the technical field of high-voltage direct-current distribution. The objective of the invention is to realize rapid and accurate positioning and distance measurement when a feeder line where a load is located has a single-pole grounding short circuit. According to the main scheme, a distance measuring module is arranged at each of the head ends of a positive feeder line and a negative feeder line to measure the distance of a fault feeder line, and the two distance measuring modules are used for measuring current waveforms to perform least square fitting to obtain respective corresponding parameters, and the parameters are substituted into a theoretical fault measuring distance formula for calculation. The method not only overcomes the defects that a traditional distance measurement method is inaccurate in measurement result and has no compatibility for a long line, but also improves the measurement accuracy, can enable a fault to be accurately positioned within the range of 40m, has remarkable superiority, and is suitable for popularization and application. A single-ended direct current power distribution network simulation model based on a modular multilevel converter (MMC) is built on MATLAB/PSCAD, and the reliability and remarkable superiority of the distance measurement method are verified.

The invention discloses an iterative filtering identification method of a linear motor transfer function, belonging to the technical field of linear motor control. The identification method in the prior art is not flexible enough and has poor adaptability, which affects the accuracy of identification parameters. An iterative learning filter is integrated into a batch-processing least squares identification algorithm in the method. A corresponding filter can be generated according to the characteristics of a system to be identified, and there is no need to design a special filter. As the filtering algorithm is added, identified data is more accurate, and the design of a high-precision controller is facilitated. The method is undoubtedly of practical significance and practical value to the design of a parameter identification and control system. The method has strong anti-jamming ability and high parameter identification precision, can help to realize high-speed and high-precision control on a system, and avoids the problem concerning filter selection and the poor anti-jamming ability of the least squares algorithm.