46results about How to "Eliminate the influence of subjective factors" patented technology

The invention discloses a multiple inspection robot cooperative operation method for a substation sequence control system. When the closed-loop control and the state visualization of the intelligent substation sequence control system are implemented, the executing and checking time of the sequence control system can be effectively saved by adopting the method. The invention provides the multiple inspection robot cooperative operation method which comprises the following steps: modeling by adopting a topological method on the basis of the substation environment, forming a topological network chart and a communication relation matrix according to walking routes of inspection robots, and carrying out search and traversal of paths to output a discharge matrix and a shortest path matrix; in the single-step sequence control process, determining a shortest communication path sequence of each inspection robot from a current position to a target point according to the distance matrix and the shortest path matrix; carrying out excellence selection of single-step sequence control, i.e. selecting the inspection robot with the optimal performance under the condition of considering comprehensive indexes of journey, time, safety and maneuverability; determining an operation path sequence of each inspection robot and a corresponding activation event; and carrying out on-line monitoring, wherein if the environment is changed, the on-line monitoring takes charge of debugging and recovering operation.

The invention relates to a transformer substation inspection robot global path planning method based on magnetic navigation. The method comprises the steps of: modeling an environment of a transformer substation to form a topology network graph and a connecting relationship matrix; determining a searching path matrix, namely the shortest path matrix, by using a Floyd algorithm, and increasing the computing speed by iterative optimization; computing the shortest connecting path between an initial point and a target point sequence by transversing nodes and edges; converting the shortest connecting path into a travel path indicated by an RFID (radio frequency identification device) sequence; and computing the rotating angle of a robot at each RFID at an intersection according to the connectivity of the path. The transformer substation inspection robot adopting the transformer substation inspection robot global path planning method provided by the invention is short in travel path, and rapid to be in place for detection, therefore, the energy of batteries is saved, and simultaneously the time for moving the robot equipment for verification after the operation of the transformer substation equipment can be effectively saved.

The invention discloses a method for identifying the mode of a switch of a substation based on infrared detection. The method comprises the following steps of: 1, inputting an image to be identified, acquiring a switch area of the image by image registration, and sequentially graying, binarizing and thinning the switch area; 2, linearly detecting the image of the thinned switch area by using a Hough transformation algorithm; and 3, determining a detection result, and identifying the state of the switch by using a switch state determination condition. An H matrix between two images is obtainedby matching of the features of scale invariant feature transform (SIFT), the switch area in the image is found, and then the state of the switch is identified. Experiments show that: by the method, the problem of identification of a power switch can be effectively solved; an important significance is provided for automatic monitoring of the power equipment of an intelligent substation; the burdens on an inspector of the substation can be reduced; and detection efficiency is greatly improved.

The invention discloses a high-accuracy detection method of real-time temperature data fusion of multiple sensors, which comprises the following steps of: step 1, remissness error removal based on relativity function sequence: (1) acquiring temperature data, (2) judging whether the number of acquisition times is larger than or equal to a set value C1 or not, if NO, continuing to acquire the temperature data, if YES, judging whether the number of the acquisition times is larger than the set value C1 or not, if NO, computing the variance and the mean value initial value of sensors, if YES, computing the variance and the mean value of the sensors in a recurrent way, (3) computing the fusion degree of the sensors, (4) computing the support degree of the sensors, and (5) selecting valid sensors, and deleting the temperature data obtained from invalid sensors; step 2, orthogonal neural network based temperature information fusion of the multiple sensors: (1) neural network training and weight vector recurrence, (2) computing neural network output, and (3) computing a temperature fusion value of the multiple sensors; and step 3, correcting variance and mean value of the invalid sensors. By the method, not only can the temperature detection accuracy and credibility be improved, related systems can be conveniently processed in real time. The method has the advantages that the stability is good, the compute is simple, the implementation is easy, and the like.

The invention relates to an engine noise source identification method. The method includes the following steps that: a noise diagnosis hierarchical analysis tree is constructed; the surface vibration signals of different components and corresponding engine surrounding noise signals are acquired through a vibration noise test; the noise signals are decomposed into different IMF components through an EEMD method, the IMF components are divided into different frequency bands according to the distribution of energy in the time-frequency spectra of the IMF components; based on the vibration signals of the different components and the IMF components of different frequency bands in the noise signals, calculation is carried out with a coherent power spectrum analysis method so as to obtain input coherent power spectra, and different input coherent power spectra are corresponding to different diesel engine components; paired comparative judgment matrixes are built according to the numerical values of the different input coherent power spectra, and the weights of the different components are calculated, and the weights of the different frequency bands are obtained; and different combined weight matrices are constructed according to the weights of the different components and the weights of the different frequency bands.

The present invention provides an online prediction method for product quality in a hydrocracking process. The method comprises: S1 based on sensitive parameters of the product in a time lag window of a process before the product quality offline test time, obtaining a steady- state index of each of the sensitive parameters by using the fitting of the polynomial least squares method; S2 based on the steady-state index of each of the sensitive parameters and the degree of influence of each of the sensitive parameters on the process, constructing the system steady-state exponent by using the Dempster rule, and according to the system steady-state exponent, obtaining steady-state data of the steady-state system in the time lag window of the process before the product quality offline test time; and S3 based on the data mean value of each of the sensitive parameters in the steady-state data in the time lag window of the process, implementing online quality prediction of the product by using the random forest product quality online prediction model. The method provided by the present invention taken full advantage of the variables in the whole hydrocracking process, so that predicted quality can reflect the overall operating condition.

The invention discloses a big-data asset assessment method. The big-data asset assessment method comprises the steps that firstly, data quality assessment is conducted, wherein data quality indexes comprise accuracy, integrality, consistency and timeliness; secondly, data size assessment is conducted, wherein data size indexes comprise the number of data attributes, the number of data tuples and the quantity of unit information; thirdly, data content assessment is conducted, wherein the data content comprises transaction data, personal information, commodity information, production management data, user evaluation data and social networking data; fourthly, industry value calculation is conducted; fifthly, data asset value calculation is conducted. The big-data asset assessment method provides a specific quantitative standard for assessment of big-data assets, the assessment process is simpler and clearer, and the subjective factor influence of assessors is eliminated, so that assessment results accord with actual conditions better.

The invention discloses an intention recognition method and device based on a convolutional neural network, equipment and a medium, is applied to the technical field of deep learning, and is used forsolving the problem of low intention recognition accuracy. The method provided by the invention comprises the steps of obtaining a target text of a to-be-identified intention; carrying out vectorization processing on the target text to obtain a target vector; inputting the target vector as an input into a pre-trained convolutional neural network, obtaining a target result vector output by the convolutional neural network, each element in the target result vector being a first probability value corresponding to each preset user intention, and the first probability value representing the probability that the target text belongs to the corresponding preset user intention; and determining the preset user intention with the highest first probability value as the target user intention corresponding to the target text.

The embodiment of the invention discloses a target object assessment and evaluation method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining an initialscore set of a target object, wherein the target object is an object to be evaluated and comprehensively scored, and the initial score set comprises initial scores of a plurality of index parametersof the target object; obtaining an objective score set of the target object through a first preset rule, wherein the objective score set comprises objective scores of a plurality of index parameters of the target object, and the first preset rule is a calculation rule representing a mapping relationship between the initial score and the objective score; and calculating a comprehensive score of thetarget object according to the objective score set. The target object to be evaluated is scored objectively and fairly, deviation of subjective scoring caused by subjective factors is corrected, thewhole process is automatically completed through the computer device, the labor cost is greatly reduced, the evaluation efficiency of the target object is improved, and meanwhile the evaluation resultis more objective and accurate.

The application provides a method and device of extracting feature vocabulary. The method comprises the steps that description information for a network resource is obtained as the extract vocabulary data; phrase marking is conducted on the extract vocabulary data to obtain the expression markup results; feature vocabulary capable of reflecting the characteristics of network resources is extracted from the expression markup results. By the adoption of the method of extracting feature vocabulary, the quantity of the extracted characteristic vocabulary can be ensured, and the quality of the characteristic vocabulary is improved.

The invention belongs to the field of medical imaging and provides a medical image parameter fitting method and device so as to improve the stability of the parameter fitting effect during medical imaging. The method includes the steps that the range of a target image in a whole medical image is determined; fitting parameters of pixels are obtained during down-sampling of the target image; the target image is sampled in the mode that the resolution ratio is gradually increased, and the fitting parameters of the pixels during up-sampling of the target image are obtained till the resolution ratio reaches the original resolution ratio of the whole medical image. Compared with the prior art that the initial value of fitting parameters of pixels is selected subjectively, the technical scheme is the mode that the initial value of the fitting parameters of the pixels is self-adaptively adjusted, manual intervention is not needed, accordingly influences of subjective factors are eliminated, fitting results are more objective and stable, and the fitting effect is more ideal.

The invention relates to a method for determining medicine compliance through an intelligent medicine box. Ideal medication time and ideal medicine dosage in the period of taking medicine are set through a man-machine interactive interface; a processor automatically records actual medication time and actual medicine dosage in the period of a man taking medicine; the above data which is stored is extracted, and a medicine compliance index of medication time and a medicine compliance index of medicine dosage are obtained through processing; a score of a user Basel evaluation scale is obtained according to the medication time and medicine dosage of the user, a difference value between the ideal medication time and the actual medication time, a difference value between the ideal medicine dosage and the actual medicine dosage and a corresponding relation between the difference values and the score of the evaluation scale are analyzed, thereby a medicine compliance classification is obtained; the processor determines a medicine compliance grade of the user, judges whether the user medicine dosage is insufficient or excessive and feeds back the information to the man-machine interactive interface. The method is conductive to achieving precise treatment of the clinical medicine.

The invention provides a distribution range determination method and device, a computer readable storage medium and electronic equipment. The method comprises the following steps: drawing a pluralityof candidate delivery ranges for a target delivery requester; estimating distribution efficiency data of each candidate distribution range according to the historical order information in the candidate distribution ranges; and determining an optimal distribution range from the plurality of candidate distribution ranges according to the distribution efficiency data of each candidate distribution range. By applying the method and the device, the distribution range can be determined more accurately, more efficiently and more objectively.

The invention discloses a seabed sand wave feature identification method, and relates to the technical field of offshore ocean engineering dynamic geologic and geomorphic environments. The method comprises the following steps: acquiring wave crest lines and wave trough lines of sand waves by utilizing actually measured water depth data and adopting an image identification method, grouping the wave crest lines and the wave trough lines of the sand waves through linear fitting and coordinate transformation, and further automatically acquiring characteristic information of the wave length, the wave height and the like of the sand waves in a large area by utilizing information of the wave crest lines and the wave trough lines of the sand waves. According to the method, the wave crest line and the wave trough line of the large-area seabed sand wave can be automatically identified, classified statistics can be automatically carried out on the wave crest line and the wave trough line of the sand wave, a large amount of sand wave characteristic data is further obtained, basic data is provided for further analyzing the relation among the wave length, the wave height, the water depth, the hydrodynamic force and the like of the sand wave, time and cost are saved, and efficiency is improved.