49results about How to "Not easy to fit" patented technology

The embodiment of the invention provides a power distribution network line loss prediction method and system, and the method comprises the steps: obtaining and cleaning the time sequence data of eachline and each transformer area in a power distribution network, employing an outlier detection method, detecting and removing the abnormal data of a time sequence, building an interpolation improved random forest model, and filling up the missing data of the time sequence; calculating the maximum mutual information coefficient of each feature and the line loss data according to the change rule ofeach time sequence feature, and selecting the feature with the maximum correlation with the line loss as the input feature of the line loss prediction model; clustering the line loss data with similarcharacteristics by adopting a k-means clustering method according to the time sequence data of the line loss of each transformer area, dividing each type of line loss data set, establishing a long-short-term memory neural network prediction model, and inputting a training sample to train the long-short-term memory neural network to obtain a line loss prediction model. The precision of short-termline loss prediction of a power distribution network can be improved, and the purpose of guiding distribution line loss management and efficiency-improving operation is achieved.

The embodiment of the invention discloses a target tracking method and a target tracking device, which are used for reducing the calculation complexity of target tracking and have stronger generalization ability. The method comprises the following steps: acquiring a first frame image to be processed from video data acquired by a camera; performing target detection on the first frame image to generate a target detection result, the target detection result including a first target detected from the first frame image; performing motion estimation on the first target by adopting a Kalman motion model to generate a target tracking result, the target tracking result comprising a prediction position of the first target in a next frame image relative to the first frame image; and estimating the motion state of the first target according to the target detection result and the target tracking result.

The invention discloses a bearing fault classification method based on CNN and Adaboost. A bearing signal is collected, the bearing signal is preprocessed, and a time domain signal and a time-frequency domain signal are extracted; a time-domain weak classification module and a time-frequency-domain weak classification module are constructed based on the time domain signal and the time-frequency domain signal; and then the time-domain weak classification module and the time-frequency-domain weak classification module are integrated and a membership probability value of a to-be-detected unmannedaerial vehicle bearing signal is predicted by using the integrated classification model. Therefore, the classification of UAV bearing faults is realized.

The invention provides a three-dimensional particle category detection method and system based on a convolutional neural network. The method comprises the following steps: constructing a three-dimensional mixed-scale dense convolutional neural network comprising a mixed-scale three-dimensional extended convolutional layer, dense connection and a loss function, training the convolutional neural network by using a three-dimensional frozen electron tomography image marked with the particle coordinates to obtain a particle selection model, and training the convolutional neural network by using thethree-dimensional frozen electron tomography image marked with the particle category to obtain a particle classification model; acquiring the three-dimensional frozen electron tomography image through a sliding window to obtain to-be-detected three-dimensional reconstructed subareas, predicting each subarea through the particle selection model, and combining prediction results of the subareas toobtain coordinates of each particle in the three-dimensional frozen electron tomography image; and extracting a three-dimensional image of each particle according to the coordinate of each particle, and inputting the three-dimensional image of each particle into the particle classification model to obtain the category of each particle.

The invention discloses a target tracking method based on triple twin hash network learning, and relates to the technical fields of computer vision, target tracking and deep learning. The method first constructs a triple twin hash network, which consists of three parts: data input layer, convolutional feature extraction layer, and hash coding layer. In the initial training process of the network, the triple twin hash network is trained using the training data set and the stochastic gradient descent backpropagation algorithm. After the training is completed, the network can obtain the initial ability of target positioning. In the tracking process, the input image is first passed through the triple-twin region recommendation network to obtain the corresponding candidate frame, and then the candidate frame is input into the triple-twin hash network for forward processing, and the similarity between each candidate frame and the query sample is calculated separately, and the highest candidate frame is selected. The candidate boxes of similarity are used as tracking target objects to achieve target tracking.

The invention discloses an integrated-learning-based indoor sound source positioning method and especially relates to features used by the sound source positioning method. Sound source signal data are converted into a feature data set by using a phase transformation generalized cross correlation function of a signal as a position feature; training and positioning testing are carried out on the feature data by using integrated learning technologies like bagging and AdaBoost; and then an integrated learning classifier is obtained and is capable of identifying a sound source position. Therefore, a defect that the performance losses are heavy in a severe environment according to the traditional sound source positioning algorithm can be overcome. Compared with the traditional positioning algorithm, the method has the following advantage: the robust sound source positioning performance can be obtained in a severe environment with high noises and reverberation in an indoor environment.