118 results about "Residual sum of squares" patented technology

The invention discloses a high-order multi-stage auto-regressive distributed lag modeling method of thermal error compensation of a numerical control machine. The method is characterized by comprising the following steps of: providing a high-order multi-stage auto-regressive distributed lag modeling formula of a thermal error of a numerical control machine containing a coefficient to be solved, selecting numerical control machine thermal error lag phases and numerical control machine temperature lag phases as 1, 2, 3 and 4 respectively, and substituting experimental data to fit the coefficient to be solved in the formula according to a least square method so as to determine high-order multi-stage auto-regressive distributed lag models of thermal error compensation of the numerical control machine in different lag phases; substituting the experimental data into each model to obtain a residual sum of squares of each model; and substituting the residual sum of squares of each model into an akaike information criterion to determine an optimal lag phase to determine a high-order multi-stage auto-regressive distributed lag model of a thermal error of the numerical control machine. The invention discloses a modeling method of thermal error compensation of the numerical control machine, which has the advantages of convenience in application, easiness in modeling and high stability, and has higher accuracy than a traditional ADL (automatic data logger) model.

The invention discloses an output-only linear time-varying structure modal parameter identification method and belongs to the technical field of structural dynamics. Firstly, a cost function of a least squares support vector machine vector time-varying autoregressive model is deduced; secondly, a function space is built by means of a Wendland compactly supported radial basis function; a regular factor is determined through the non-parameter method based on Gamma testing, and a basis function width reduction coefficient is given on the basis of actual experiences; a time-varying autoregressive model order is determined according to the Bayesian information criterion and the Akaike information criterion; a function space order is determined according to the ratio of residual sum of squares to sequence sum of squares; finally, the matrix expression of the least squares support vector machine vector time-varying autoregressive model is solved according to the cost function, modal frequency of a system is solved according to a time freezing method, and linear time-varying structure modal parameter identification is finished. The method can improve calculation efficiency, improves system robustness, and is widely used in linear time-varying structure modal identification in structural dynamic engineering application.

The method for forming the optimal characteristic curve of a solar cell comprises the steps of: providing a first acceptable error; and determining a current-voltage polynomial regression equation, whose square root of the residual sum of square is less than the first acceptable error for a set of solar cell measured data. The order of the current-voltage polynomial regression equation is gradually increased until the square root of the residual sum of square of the current-voltage polynomial regression equation is less than the first acceptable error.

The invention relates to an equivocation determining method, in particular to a dynamic equivocation determining method. The invention adopts an LAMBDA method to search a plurality of groups of equivocations, and then utilizes the square sum of residuals which is alternatively combined to detect, thereby the reliability of the equivocation determination can be increased. The invention utilizes the known wide-lane equivocation and L1 equivocation to express the non-ionosphere equivocation combination for carrying out the equivocation search so that the influence on a first-order phase of an ionosphere is eliminated, and the integer performance of the searched equivocation can be caused, thereby the distance of the equivocation search is enlarged, various conditions for rapidly and dynamically determining the long-distance equivocation are fully utilized, and the equivocation search is accelerated.

The invention discloses an optimum eight-place standardization method of a flexible gyroscope, the flexible gyroscope is arranged on a three-shaft place rate rotary table, wherein, the flexible gyroscope is connected with data collecting equipment which is communicated with a computer. The invention discloses a specific position in the optimum eight-place standardization, and the output of the flexible gyroscope is effectively improved in the way that the invention compensates a measurement value of a compensation model G0 of obtained optimum eight-place drifting coefficients and vibrating reed gyro quiet errors. In a testing process of the vibrating reed gyro in an inertial navigation testing center, a traditional eight-place method and the optimum eight-place method are respectively adopted to obtain the drifting coefficients which respectively and objectively output compensated evaluation results of another three-place top in space. By residual sum of square of a top measurement value, compared with the traditional eight-place method, the results of compensated drifting coefficients solved by the optimum eight-place experiment design method of the flexible gyroscope are increased by 4 to 8 times.

The invention discloses a method for forecasting the acute toxicity of organic compounds by building a quantitative structure-activity relationship model with a quantum chemistry method. The method fully geometrically optimizes compound structures by using a Gaussian procedure so as to obtain quantum chemistry parameters including molecular volume, relative molecular mass, highest occupied molecular orbital energy, lowest unoccupied molecular orbital energy, energy gaps of frontier molecular orbital, dipole moment, solvation energy, electron energy and the like; using the quantum chemistry parameters and a hydrophobicity parameter as structural descriptors; in combination with toxicity data, quantitative relationship equations between various structural descriptors and toxicity are established according to a written procedure based on partial least square stepwise linear regression to obtain the multiple correlation coefficient, F-test value and sum of squared residuals, and then the model is verified so as to guarantee the external predictive ability. Therefore, the method can quickly and effectively forecast the toxicity of organic compounds to be studied, and provide necessary basic data for risk assessment and supervision of chemicals.

An optimum measurement subset with a specified number of elements is generated from a set of input global navigation satellite system (GNSS) measurements. A design matrix and a weight matrix are generated. Values of a set of coefficients corresponding to the set of input GNSS measurements are calculated. The value of a specific coefficient is calculated as the ratio of the change in value of the at least one target parameter resulting from the change in value of the specific input GNSS measurement to the change in value of the sum of squared residuals resulting from the change in value of the specific input GNSS measurement. The optimum measurement subset is selected based at least in part on the values of the set of coefficients.

The invention discloses a method for forecasting data. The method comprises the following steps of: acquiring a trend value and a residual value of each observation point; determining the principal direction of data forecasting; calculating the distances between one observation point and other observation points one by one, and included angles between the connection lines between one observation point and other observation points and the principal direction; classifying the observation points; calculating the average residual sum of squares of each type of the observation points respectively according to the residual values respectively; and calculating the distance between a target point and each observation point and an included angle between the connection line between the target point and each observation point and the principal direction, selecting the average residual sums of squares corresponding to the distances and the included angles, and forecasting the observed value of the target point. By applying the technical scheme, the distance correlation and the direction correlation are considered comprehensively in the data forecasting process, the accuracy of the average residual sum of squares is improved relative to Kriging interpolation in which only the distance correlation is considered, so that accuracy of the observed value forecasting of the target point is improved.

Provided is a cable running state assessment method and system. The method is characterized by obtaining characteristic quantity of a cable, wherein the characteristic quantity comprises insulation resistance value, insulation resistance absorption ratio, average value of dielectric loss angle tangent values, linear fitting slope, residual sum of squares, partial discharge quantity, partial discharge quantity increment, partial discharge frequency, concentrated partial discharge point positions and harmful discharge information; judging grade of each characteristic quantity according to a preset standard; and obtaining a state assessment result according to the weight corresponding to the grade of each characteristic quantity. According to the cable running state assessment method and system, the obtained characteristic quantity comprises the insulation resistance value, the insulation resistance absorption ratio, the average value of dielectric loss angle tangent values, the linear fitting slope, the residual sum of squares, the partial discharge quantity, the partial discharge quantity increment, the partial discharge frequency, the concentrated partial discharge point positions and the harmful discharge information, so that the running state of the cable cam be evaluated more completely and accurately.

The invention provides a method and device for predicting the output of an oil and gas well, and the method comprises the steps: obtaining the actual effective output data of the oil and gas well in a plurality of historical production time periods; respectively predicting the output of the oil and gas well in at least one historical production time period and one future production time period through employing a hyperbolic decline method, an exponential decrease method and a curve fitting method and the actual effective output data of the oil and gas well; calculating the residual sums of squares of prediction results of the above three methods according to various types of obtained output prediction data of the oil and gas well and the actual effective output data of the oil and gas well; selecting the minimum residual sum of squares, and enabling the data, obtained through the method which obtains the minimum residual sum of squares, of the oil and gas well in the future production time period to serve as the final output prediction result of the oil and gas well. The method selects the oil and gas output prediction result with the highest accuracy from the results of three prediction methods through residual error evaluation.

The invention relates to an adaptive data driving fault diagnosis method and a device in a complex refining process. The method comprises steps that, a relation graph among relevant process parameters causing a fault is established, and the process parameters causing the fault are determined; time sequence of all the process parameters and time sequence of the fault are acquired through the process parameters; whether the time sequence of the process parameters is covariance stationary is determined; if not, first order difference processing on the time sequence of the process parameters is carried out; a regression equation is constructed by utilizing the time sequence of the process parameters and the time sequence of the fault after covariance stationary, and a first residual error quadratic sum and a second residual error quadratic sum are acquired according to the regression equation; F examination is carried out; a largest path of a cause and effect relation magnitude is searched according to the relation graph; the time sequence of the process parameter which corresponds to a start point of the path is the root reason of the fault, and the terminal point of the path corresponds to the time sequence of the fault; and the largest path of the cause and effect relation magnitude is the propagation path of the fault in a system.

The invention discloses a Fresnel matrix THz wave propagation model based high-precision thickness detection method for coatings. Based on a electromagnetic field theory model, a reflecting propagation one-dimensional electric field model of normal incidence THz wave in a multi-layer medium structure is established according to a reflecting type THz time domain spectrum system used in coating thickness detection. On this basis, a Fresnel coefficient matrix and a phase matrix are introduced and a Fresnel matrix THz wave propagation model is proposed and validated. Quantitative detection of coating thickness is realized by utilizing the Fresnel matrix THz wave propagation model based LS (Least Squares) thickness optimization method and the coating thickness is solved through iterative optimization so as to minimize the minimal residual sum of squares of a simulation result and a measurement result and to realize the high precision optimization of thickness parameters. A distance parameter DTS between a THz emitter and a to-be-measured piece is introduced and visible treatment is performed on distance errors between the THz emitter and the to-be-measured piece, so that the accuracy ofcoating thickness detection is improved further.

The invention discloses a multiple linear regression model-based belt weigher main error factor analysis method. The method includes the following steps that: a belt weigher experiment platform is set up, main error factors such as tensions, temperatures and equivalent flow rates are changed according to actual situations, the tensions, temperatures, equivalent flow rates, calibration values and hopper weighers of various test points are recorded, and the relative errors of the calibration values and hopper weighers are calculated; the correlation coefficients of the main error factors of a belt weigher are calculated, and the correlation of the main error factors is judged; the relations of the factors are determined according to the correlation, and a multiple linear model is set and solved; residual sum of squares, determination coefficients and MS variance of residuals are adopted as check indexes, and the regression effect of the established model is evaluated; and the test values of the test points are predicted through using a fitting model, and the test values are compared with practical values, so that prediction errors can be calculated, and the accuracy of a fitting result is determined. With the method of the invention adopted, a theoretical basis is provided for quantitative analysis on the degree of influence of the main error factors on the accuracy of the belt weigher.

The invention provides an estimation method of external network Thevenin equivalent parameters based on single point measurement information, relating to the power system external network Thevenin equivalent parameters estimation method technology field. According to the invention, a computer and a data acquisition apparatus of a line terminal are utilized, through a program, firstly a parameter of a monitored line is inputted, then according to SCADA measurement data of a plurality of periods of a terminal of the monitored line, a least square equivalence model with minimum of square sum of residuals as an object is established, the model is solved, and equivalent built-in potential and an equivalent impedance parameter of a network outside of the monitored line are obtained. According to the invention, a physical meaning of the external network Thevenin equivalent parameters is fully considered, rationality of the external network equivalent parameters can be effectively ensured, influence of a measurement error on an equivalent parameters calculation result can be effectively avoided, accuracy is high, the method is simple, estimation precision is high, a calculation speed is fast, and the method can be widely applied to the estimation of the external network Thevenin equivalent parameters of any power line whose terminal is installed with the data acquisition apparatus.

The invention provides a method for confirming yield through spectrum unfolding analysis of a stratum element logging instrument. The method comprises the following steps of confirming a segment area number n and left and right boundaries kn and kn+1 of each segment area; calculating a standard trap spectrum normalization response matrix A; calculating a total numerical value ci of a spectrum to be unfolded in each segment area; calculating an opposite angle weighting matrix W; calculating a weighted least square method spectrum unfolding to obtain the yield yj; utilizing a mean value delta 2 of residual sum of squares to judge an imitative effect. The invention adopts a segment clamping manner spectrum unfolding method. The segment clamping manner is characterized by selectively selecting a channel area, in which an characteristic energy spectrum of an element is located, to unfold the spectrum. Through MC numerical simulation, the method has no large effect on spectrum unfolding.

The invention discloses an evaluation method and system of water quality based on fish electrocardio indexes. The method comprises the following steps: receiving fish body electrocardio signals beforeand after fish contact with a solution which contains specific organic matter, and respectively extracting multiple kinds of electrocardio indexes; analyzing the relativity between environmental stress and each extracted electrocardio index according to a Pearson correlational analysis method, and screening the electrocardio indexes for evaluating the specific organic matter in the water quality;analyzing and evaluating environmental stress and linear regression models between various electrocardio indexes according to Akaike information criteria and the residual sum of squares so as to obtain the electrocardio indexes used for finally evaluating the organic matter in the water quality. Receive the fish body electrocardio signals before and after the fish contact with to-be-determined water, and respectively extracting an electrocardio index which is corresponding to a certain organic matter in determined evaluation water quality; contrast the extracted electrocardio indexes of the fish body electrocardio signals before and after the fish contact with the to-be-determined water so as to obtain electrocardio index changes for evaluating the corresponding organic matter in the to-be-determined water quality.

The invention provides a method for determining the yield of a stratum element logging instrument by virtue of spectrum unfolding. The method comprises the following steps: calculating a standard capture spectrum normalization response matrix A, calculating actual measurement stratum capture spectrum count ci, calculating an opposite angle weighting matrix W, calculating the yield yj obtained by virtue of spectrum unfolding based on a weighted least squares method, and judging the fitting effect by use of residual sum of squares and the average number delta 2. According to the method, a one-by-one spectrum unfolding method is adopted, complete 256 energy windows are adopted for spectrum unfolding. According to MC numerical simulation, the method has slight influence on spectrum unfolding; a one-by-one method is adopted for performing spectrum unfolding on an unknown stratum, resolution result errors are large but can basically reflect the types and relative contents of elements in the stratum.

The invention discloses a Spectralon diffuse reflection plate correcting method. The method comprises the steps: 1) illuminating the Spectralon diffuse reflection plate from different zenith angles by a light source, and using the spectrograph to collect the corresponding reflection spectroscopic data from the different zenith angles and the azimuth angles; 2) intercepting the spectroscopic data within the valid wave band range from the reflection spectroscopic data, calculating the average value, and eliminating the spectroscopic data with relatively large difference; 3) calculating the residual sum of squares Rss of the residual spectroscopic data, determining the value of the most stable probe receiving zenith angle (thetar, degree), and establishing the relational expression of the random probe receiving zenith angle thetar and (thetar, degree); and 4) substituting the relational expression to the BRDF expression of the Spectralon diffuse reflection plate, and finishing the Spectralon diffuse reflection plate correction. The method is capable of using the known data of the Spectralon diffuse reflection plate to correct, calculating the universal expression of the probe receiving zenith angle orientation, and shortening the time for the follow-up test process.

An LS (least square)-based multi- model adaptive state estimation method includes the following steps: S1, same moving object observation sequences of sensors are acquired to form moving object X, Y and Z direction observation sequences in the space rectangular coordinate system at different times; S2, the degrees and the number of time-variable polynomial equations are determined, and LS (least square)-based polynomial equation coefficient estimation of the moving object X, Y and Z direction observation sequences is performed respectively; S3, according to the polynomial equations in the step S2, regression sum of squares and residual sum of squares are respectively calculated for F-test to determine a best fit polynomial; S4, a corrected best fit polynomial is determined; S5, a first time partial derivative and a second time partial derivative are sought respectively to obtain moving object corresponding velocity and acceleration polynomials; and S6, object position state estimator coordinate conversion is performed to obtain corresponding position coordinates, and then by combining with velocity estimated values, object state estimation can be completed.

The invention discloses a precision calibration method of a wide-field infrared photoelectric theodolite, relates to the field of photoelectric measuring and controlling, and solves the problem in the prior art that an error is led to the processing and debugging processes and is not completely consistent with an optical design value, resulting in low photoelectric measurement precision of the wide-field infrared photoelectric theodolite. The precision calibration method comprises the following steps: calibrating a detection rack; collecting model data; building a calibration model and precision detection; obtaining a root-mean-square value of a measurement angle. A multi-element regression analysis method is applied to correction of a miss distance, so that the residual sum of squares between an observation value and a regression value is the minimal, and the regression precision is improved. The wide-field is subjected to partition processing, multi-element regression analysis calculation is carried out on each subarea, and improvement of the calculation precision is facilitated, so as to indirectly improve the measurement angle precision.