86results about How to "Low accuracy requirements" patented technology

The invention discloses an electric vehicle wireless charging system with metal foreign substance detection. The system comprises a wireless electric energy emitting device, a wireless electric energy receiving device, and a metal detection device, wherein the input end of the wireless electric energy emitting device is connected with a power supply component, the output end of the wireless electric energy emitting device is connected with a wireless electric energy emitting coil, and the wireless electric energy emitting coil is capable of producing a high frequency resonant magnetic field with fixed frequency; the input end of the wireless electric energy receiving device is connected with a wireless electric energy receiving coil, and the wireless electric energy receiving coil is coupled with the wireless electric energy emitting coil; when the resonant frequency of the wireless electric energy receiving coil is the same as the fixed frequency produced by the wireless electric energy emitting coil, the wireless electric energy receiving coil end produces induced electromotive force and is used for charging an accumulator of the electric vehicle connected with the output end of the wireless electric energy receiving device; the metal detection device comprises a balance structure metal detection coil laid on the surface of the wireless electric energy emitting coil; the metal detection coil and a metal detection circuit form a loop to detect whether a metal foreign substance is existent in the magnetic field.

The method discloses a method for sorting signals of a radar radiation source by using coverage in a complex dense environment and relates to a radar radiation source signal sorting method by using coverage, aiming at solving the problem of poor sorting ability of the prior art in the complex dense environment as well as in a serious overlapping condition of characteristic parameters. The method comprises the following steps of: choosing the characteristic parameters of a radar signal capable of being standardized as input characteristic vectors, initializing a radar signal sample parameter space; carrying out supervisory learning on the radar signal with parameter prior information in the radar signal sample parameter space by using a coverage algorithm; separating the radar signals of different kinds by using different fields by using the coverage algorithm; expressing the membership degree of the characteristic parameters of each radar signal to be sorted and identified in relative with each field by using a cloud model, and obtaining the membership value of the membership degree of the radar signal to be sorted and identified in relative with each field; and sorting the signal to be detected by using decision rules. The method for sorting signals of the radar radiation source by using coverage in the complex dense environment is suitable for sorting radar radiation source signals.

The invention discloses a wireless charging, alignment and metal foreign body detection system of an electric automobile and a method thereof. The detection system comprises a transmitting coil, a receiving coil, a probe, a signal conditioning module and a control module connected with the signal conditioning module, wherein the transmitting coil is connected with the output end of a wireless charging and transmitting device on the ground surface of a parking lot, the receiving coil is connected with the input end of a wireless electric energy receiving device of an electric automobile chassis and is mutually coupled with the transmitting coil, and the probe is laid on the surface of the transmitting coil and is mutually coupled with the transmitting coil and the receiving coil. The probe is of an axis-symmetric structure and is composed of two coils equal in size and turn number and opposite in winding direction. The probe is consistent with the boundary of the transmitting coil in size. The signal conditioning module is used for receiving voltage signals output by two output ends of the probe and conducting amplification, filtration and peak-holding processing on the two signals output by the two output ends. The control module is used for obtaining the difference of the two received voltage signals and comparing the obtained voltage signal difference value with a preset voltage threshold value.

The invention discloses a fuzzy sliding mode and neural network based composite control system of a permanent synchronous motor. The control system comprises a rotation speed control outer ring and a current control inner ring, wherein a control unit in the rotation speed control outer ring adopts a composite control module; the composite control module comprises a fuzzy sliding mode control module and a neural network control module; and the fuzzy sliding mode control module and the neural network control module are combined together through two weighing factors lambda n and lambda f. The control system has the advantages of both the fuzzy sliding mode control module and the neural network control module, so that the strong robustness and high stability of the control system of the permanent synchronous motor are realized, the steady state precision is improved, and moreover, the rapidness in control and the strong load disturbance resistance are achieved.

The invention provides a fractional order based control system and a fractional order based control method for direct torque of a permanent-magnet synchronous motor for an electric vehicle, so as to effectively adjust a speed loop of the control system for the direct torque of the permanent-magnet synchronous motor. The control method includes the steps: firstly, detecting three-phase current, voltage and inverter direct-current bus voltage which are outputted by an inverter, obtaining electromagnetic torque, stator flux linkage and a stator flux linkage sector theta at the current state by torque, flux linkage and theta calculation units, comparing the calculated flux linkage and torque with a given flux linkage and given torque signals outputted by a fractional order adjuster respectively, outputting the current flux linkage error and torque error states, determining a selected voltage vector by the states and the sector of the flux linkage together to determine on-off state of the inverter so that the permanent-magnet synchronous motor is controlled, wherein the speed loop adopts fractional order PIepsilon Dmu control. The control system has the advantages of simple structure, novel conception, rapidness in response and the like, and the control method has the advantages of wide control range, high precision, rapidness in response and the like as the speed loop adopts fractional order PIepsilon Dmu control.

The invention provides an identification method for illegal cooking oil, and mainly solves the problems in the present identification method for illegal cooking oil. The identification method disclosed by the invention is mainly that: the mass fraction for palmitoleic acid occupying the total amount of fatty acid in grease is detected to identify whether the oil is illegal cooking oil or not; and in the practical detection, a sample is detected by adopting a gas chromatograph-mass spectrometer or a gas chromatography after being saponated and performed with methyl esterification, so as to identify whether the oil is illegal cooking oil or not after adopting peak area or peak height normalization quantitation. The method has the characteristics of definite target spot, strong feature, simplicity in operation, strong repeatability, application in quick detection and the like.

The invention discloses a moving vehicle speed measurement method based on target tracking and feature point matching, and belongs to the field of intelligent traffic. The method mainly comprises the steps that a current image sequence Fn is obtained; a gray level image Y(n) of the current frame is obtained; position information P_Cur of a moving vehicle of the current frame is obtained according to background modeling; track information T of the moving target of the current frame is obtained; feature points are extracted, and information of the feature points is recorded; the feature points are matched, the moving displacement of the target is calculated, and the instant running speed of the target is obtained according to a camera calibration result; scale changes of feature points to be matched and the matched feature points are calculated according to a final matching result and then are analyzed, and algorithm parameters are updated according to the analyzing result; the average value of all instant speeds of the moving target is obtained. The method can efficiently and accurately calculate out the moving displacement of the vehicle and is applicable to the systems such as security vehicle monitoring systems and red light running electronic police systems.

The embodiments of theinvention provide a substrate and a manufacturing method and display panel thereof, and relates to the technical field of display. The substrate, and the manufacturing and display panel thereof allow even spreading of ink drops in the grooves that are defined by pixel defining layer so as to effectively increase product quality and display effects. The substrate includes a substrate and a pixel defining layer which is disposed on the substrate. The pixel defining layer includes a dam body and a plurality of micrometer-protrusions which are disposed on the supper surface and the lateral of the dam body. The protrusions have lyophobic properties.

The invention provides a taxi call center system and an implementation method thereof. The system comprises a client communication terminal, a taxi driver communication terminal and an intelligent network platform, wherein the client communication terminal is used for allowing a client expecting to use a taxi service to dial an access number to provide the coordinate information of the prior position and telephone number information to the intelligent network platform; the taxi driver communication terminal is used for allowing an idle taxi driver to provide the coordinate information of the prior position and telephone number information to the intelligent network platform by dialing the access number; and the intelligent network platform is used for recording the coordinate information of the prior position and the telephone number information respectively which are provided by the client communication terminal and the taxi driver communication terminal, and performing matching, wherein if the intelligent network platform finds a matched record, the intelligent network platform puts through the call, and otherwise, the intelligent network platform continues to find out the matched record. The taxi call center system and the implementation method thereof can effectively improve the convenience and the practicability of calling a taxi and reduce the construction and operation cost of a taxi call center.

The invention discloses a mounting structure and a mounting method of a decorative plate. A keel of the mounting structure of the decorative plate is fixed on a wall body, two sides of the lateral wall II of a decorative plate edge hanging piece are hung on the wall of an opening groove of the keel, the lateral wall I of the decorative plate edge hanging piece is hooked on a groove of the decorative plate, and the length direction of the keel is perpendicular to the length direction of the corresponding groove on the decorative plate. The method comprises the following steps of: (1) paying off; (2) fixing the keel; (3) mounting the decorative plate edge hanging piece on the keel; (4) fixing the decorative plate. The edge of the decorative plate and wall body are fixed through the keel and the decorative plate edge hanging piece, and when the edge hanging piece is mounted, the groove fixed on the decorative plate can be utilized, thereby being convenient to mount and easy to implement.

The invention discloses a SAWR structure with a transverse mode suppression effect. By virtue of a surface acoustic wave resonator utilizing the structure, a transverse mode of the SAWR can be effectively suppressed through the optimal design of structural parameters, so that very high quality factor is realized. A resonator structure for the high-performance surface acoustic wave filter comprisesa piezoelectric monocrystal substrate, an interdigital electrode IDT, a reflecting grating, a busbar and a Si-base perfect matching layer, wherein the interdigital electrode IDT mainly contains an Alor Cu metal material, and the busbar mainly contains the Al or Cu metal material; the standard thickness h of the IDT and a reflecting grating electrode is more than or equal to 0.02lambda and less than or equal to 0.2lambda; and the value range of an aperture w is more than or equal to 20lambda and less than or equal to 50lambda, the value range of a false point length l is more than or equal to2lambda and less than or equal to 10lambda, the value range of an aperture gap g is more than or equal to 0.2lambda and less than or equal to 0.5lambda, the value range of a metallization ratio r ismore than or equal to 0.4lambda and less than or equal to 0.8lambda, and lambda is a resonant wavelength. According to the SAWR structure provided by the invention, by adding a false point and adjusting the length of the false point, the transverse mode interference of the SAWR can be effectively suppressed.

The method includes steps as the follow. Weight value P1 of first proof weight is weighed out through the present balance and to stove it in microcomputer; weight value P2 of a standard weight having the same nominal value with the first proof weight is weighed out by the same balance and to store it in microcomputer; D-value is calculated out with the microcomputer and D=P2-P1 is used as correction for the first proof weight; D is led in when calibrating is carried on to obtain correct calibration coefficient K and to store it in microcomputer.

The invention, which belongs to the field of the closed-loop control technology of the permanent magnet synchronous motor, relates to a method for determining the rotational speed of a two-phase orthogonal-winding permanent magnet synchronous motor, especially to a method for determining a rotational speed of a two-phase orthogonal-winding permanent magnet synchronous motor on the condition of having no speed sensor, particular to a method for counter EMF observation of the two-phase orthogonal-winding permanent magnet synchronous motor and rotational speed determination. The method provided by the invention can be applied to a speed-sensor-free closed-loop control system of a two-phase orthogonal-winding permanent magnet synchronous motor so as to solve problems of low system reliability, poor environment adaptability, and increasing cost caused by sensor installation at the motor control system.

The invention discloses an attitude estimation method of a three-axis stabilized satellite without a gyroscope. The method comprises the steps of estimating an inertial angle velocity of the satellite by using a satellite attitude dynamics equation, thereby replacing inertial reference units of measuring the angle velocity of the satellite; carrying out filtering estimation on an attitude determining value by using a satellite kinematics equation, a star sensor, an infrared earth sensor and a digital sun sensor; and carrying out satellite attitude angle estimation by using attitude determining angles of the star sensor, the infrared earth sensor and the digital sun sensor and taking an attitude measurement value of the star sensor as a benchmark. The beneficial effect that different inertial reference units carry out satellite attitude angle and attitude angle velocity estimation by adopting the star sensor, the infrared earth sensor and the digital sun sensor to improve the reliability of a satellite attitude and orbit control system is obtained.

The invention discloses a robust STAP method based on array manifold priori knowledge having measuring error, and the method comprises the steps: S1, a clutter space-time steering vector set is acquired according to a given error range; S2, an important space-time steering vector is selected in the clutter space-time steering vector set, and an eigenvalue and an eigenvector of the important space-time steering vector is calculated; and S3, a clutter covariance matrix is obtained according to the eigenvalue and the eigenvector of the important space-time steering vector, and a filter weight vector is obtained according to the clutter covariance matrix. In the method, errors are unavoidably existed in the acquisition of array manifold knowledge and directly causes processing performance limitation of STAP based on the array manifold knowledge. Compared with the STAP method based on the array manifold knowledge in the prior art, the method provided by the invention reduces the requirement for the accuracy of priori knowledge, has robust characteristic for the priori knowledge with certain errors, and can avoid a process of clutter covariance matrix inversion in a process of designing a filter, so that a purpose of reducing system computation complexity can be achieved.

The invention discloses a method for storing business documents and a system. The method for storing business documents includes acquiring a service document, and judging whether the service document is an initialized document or not; acquiring the date of the initiated document if the service document is the initiated document, comparing the date with an enabled time of the system, storing the initiated document if the date of the document is earlier than the enabled time of the system, and abandoning the initiated document if the date of the document is later than the enabled time of the system; and storing the service document to a balance statement if the service document is not the initiated document. The technical scheme is applied, when the acquired service document is the initiated document, the date of the initiated document is compared with the enabled time of the system, and the initiated document earlier than the enabled time of the system is stored. Accordingly, when recording of the initiated documents makes mistakes, the complete initiated documents do not need to be recorded again, implementation cost is saved, and workload of operators is reduced. Simultaneously, initiation preparation work for the system is simplified, and online time is shortened.

The invention discloses a bill scanning image fast tilt correction method and system, and the method comprises the steps: step 1, obtaining a color scanning image of a bill, and sequentially carryingout the graying and binarization processing; step 2, performing image pyramid processing on the binary image obtained in the step 1 to obtain a plurality of binary images with different resolutions; step 3, selecting a binary image on the top layer of the image pyramid as a rough detection object to perform inclination angle rough detection to obtain a rough detection inclination angle; step 4, based on the rough detection inclination angle, selecting an image pyramid non-top binary image as a fine detection object to carry out inclination angle fine detection to obtain a fine detection inclination angle; and step 5, carrying out rotation correction processing on the bill scanning image according to the fine detection inclination angle. According to the invention, while the inclination angle detection precision is ensured, the inclination angle detection speed can be accelerated, so that the method has relatively high real-time performance and detection precision.

The invention discloses an automatic guitar polishing device. It includes a box body, a first roller is arranged inside the upper end of the box body, horizontal connecting rods are connected to both sides of the first roller, a vertical sliding rod is connected to the middle of the horizontal connecting rod, and a fixed rod is connected to the other end of the vertical sliding rod. The vertical slide bar is connected with a slide block through a chute, the slide block is connected with the fixed bar through a spring, and a third roller is connected to the inner side of the slide block on the two vertical slide bars; a guitar mold is arranged below the horizontal connecting bar, Both sides of the bottom of the guitar mold are provided with a rack, the rack is connected with a gear, and the gear is connected with a second motor; the first, second and third drums are wound with abrasive belts. The invention has the advantages that the abrasive belt is not easy to slip, the installation of the abrasive belt is simple, the accuracy of the length of the abrasive belt is low, the service life of the abrasive belt is long, the degree of automation is high, the manual grinding environment is good, the grinding flatness is high, and the workload is small. Beneficial effect of high sanding work efficiency.

The invention discloses a load-compensation-based large-diameter pile foundation self-balanced detecting method and belongs to the technical field of constructional engineering detection and evaluation. According to the method, firstly, collection of geological survey data in a test pile position is conducted, and arrangement and installation position of a test pile load box are determined initially; then an upper loading face and a lower loading face of the load box and the pile top and the pile end of a test pipe are all provided with steel bar stress strain sensors and displacement meter sensors, a counter-force loading device is arranged at the pipe top of the test pile to provide compensative load possibly needed in a pile foundation test, according to geological survey data, through pile foundation monitoring technology and a load-compensation system, an integrated effective load-compensation-based large-diameter pile foundation self-balanced detecting method is formed. The method is suitable for pile foundation detecting of bridge structures including high-rise or super high-rise or tall building structure pile foundations, or large span continuous beams or large span continuous rigid frame bridges or cable-stayed bridges or suspension bridges.

The invention discloses a steel bar connecting device for a precast concrete component and a steel bar connecting method for the precast concrete component. The connecting device comprises a hoop and a plurality of connecting screw bolts, wherein the hoop is tightly hooped at a connecting part between two steel bars to be connected; the connecting screw bolts are uniformly arranged on the hoop; the hoop comprises an outer hoop body sleeving on two steel bars and three connecting plates arranged at the outer side of the outer hoop body; each outer hoop body is in a 3 shape and is formed by connecting two arc-shaped plates; a plurality of connecting screw bolts are arranged on the three connecting plates; a plurality of screw bolt installing holes through which a plurality of connecting screw bolts are installed are formed in the three connecting plates; the three connecting plates are in parallel arrangement and comprise a first connecting plate, a second connecting plate and a third connecting plate; and two steel bars are tightly hooped between the two arc-shaped plates and the third connecting plate. The connecting method comprises the following steps of: 1, steel bar penetration; and 2, screw bolt tightening. The connecting device and the connecting method have the advantages that the design is reasonable; the use and the operation are simple and convenient; economic performance and practicability are realized; the steel bar connecting process can be simply, conveniently and fast completed; and the connection is reliable.

The invention provides an electromagnetic grabbing device for picking up iron magnetic objects, and belongs to the field of mechanical devices. The electromagnetic grabbing device comprises a base, a cable, columnar electromagnets, blocking pieces and clamping pieces. A plurality of fixing through holes are formed in the base. The blocking pieces are arranged in the fixing through holes. Conical holes with diameters gradually decreased from top to bottom are formed in the blocking pieces. Conical surfaces with diameters gradually decreased from top to bottom are arranged on the outer surfaces of the clamping pieces. The clamping pieces are inserted in the conical holes. The clamping pieces and the blocking pieces are provided with central through holes. The columnar electromagnets are inserted in the central through holes. The cable, on one hand, is fixedly connected with the base, and on the other hand, is electrically connected with the columnar electromagnets. By means of the electromagnetic grabbing device, even when the upper surface of a falling object is of a non-planar structure, the bottoms of the columnar electromagnets can act on the upper surface of the falling object, the contact area between the columnar electromagnets and the falling object is large, and large magnetic attraction force can be provided after power-on; thus, falling objects complex in shape and large in mass can be picked up.

The invention discloses an instrument reading neural network identification method, which belongs to the field of visual identification and automation, and comprises the following steps of: (1) acquiring a pointer instrument image: shooting and recording the pointer instrument image; (2) acquiring a correction reference; (3) training a neural network of instrument panel images at different positions and a digital display reading result, and learning parameters by applying a feedforward algorithm; and (4) performing instrument reading identification detection: inputting a pointer instrument camera image, and giving a corresponding instrument reading identification result and a confidence interval by a neural network model. The defects of an existing instrument reading extraction method areovercome, the parameter learning capacity of the artificial neural network model, the learning capacity based on the artificial neural network model and the physical significance of error feedforwardare fully utilized, the pointer instrument reading is accurately detected by processing the shot image, and the invention has the advantages of being high in real-time performance, high in accuracy and the like. Good practical values are realized.

The invention discloses an elevator vibration attenuation method and device. The elevator vibration attenuation method comprises the steps that the current actual elevator car load and the current elevator floor position are obtained; a car load is selected from recorded car loads, wherein the difference between the selected car load and the actual elevator car load is within the preset range, and the selected car load is used as the current car load; according to the corresponding relation among recorded car loads, the elevator floor position and servo motor rotation displacement, the servo motor rotation displacement corresponding to the current car load and the current elevator floor position is determined; according to the determined rotation displacement, a servo motor installed on a guide shoe is adjusted. According to the elevator vibration attenuation method and device, the requirement for acceleration signal collection real-time performance and accuracy is lowered, the servo motor adjusting accuracy and the elevator riding comfort level are improved, and meanwhile the problem that the guide shoe is damaged caused by over adjustment of the servo motor is avoided due to the fact that the rotation displacement of the servo motor is a known and determined adjustment variable.

The invention discloses an intelligent TV (television) and a channel input method thereof based on a touch remote controller. The method comprises the following steps that: after the intelligent TV confirms that a channel input command is received, channel numbers with single bits, which are annularly arranged in sequence, are displayed; one of the displayed channel numbers is selected as a selected channel number with a single bit, and the selected channel number is displayed in highlight; and if the intelligent TV confirms that channel input is finished, a channel corresponding to the selected channel number is switched to. When a user requires to switch channels, a ring of annularly arranged channel numbers are displayed on a display screen, the finger of the user can slide on a touch panel of the touch remote controller to select a channel number in a sliding way; and compared with that of the prior art, the requirements on the operation precision and accuracy of the user are greatly reduced, so that the channel input operation of the user can be facilitated.

The invention discloses a blood return prevention time-controllable filter for a medical infusion apparatus and the medical infusion apparatus. The filter comprises a barrel body, a water storage tank and an infusion passage, wherein the water storage tank is arranged in the barrel body, and is used for storing residual liquor for preventing blood return; the infusion passage is used for conventional infusion; the water storage tank has a first or second structure; according to the first structure, at least one through hole is formed in the water storage tank, and is provided with a flow limiting passage, so that the residual liquor drips through the flow limiting passage at blood return prevention speed; according to the second structure, at least one through hole is formed in the water storage tank, and a flow limiting tube is connected to the interior of each through hole in a clamping way, and is provided with a hollow flow limiting passage, so that the residual liquor drips through the flow limiting passage at the blood return prevention speed. According to the filter and the medical infusion apparatus, enough time for the slow dripping of the liquor in the water storage tank through the flow limiting passages to overcome blood spreading venous pressure at the end of infusion of the medical infusion apparatus can be ensured, and sufficient time for the prevention of blood return during infusion can further be ensured.

The invention discloses a method for measuring low-range COD in a water body by a dual wavelength internal standard method. The method involves the following steps: adding a mixed solution of sulfuricacid and phosphoric acid of potassium dichromate and copper ions to a water sample, digesting at temperature of (150-165) DEG C for (10-15) min, selecting a measurement wavelength at a wavelength of(335-460) nm, and selecting a calibration wavelength at (750-900) min. The method for measuring a change in wavelength absorbance after calibration and for measuring the COD in a range of (0-250) mg / Lis established. According to the method, requirements for consistency and accuracy of oxidant addition amount for measuring the low-range COD are reduced, and the accuracy of low-range COD measurement is improved, especially technical indexes such as the measurement accuracy of the low-range measurement.

The invention discloses a ship trajectory prediction method based on machine learning and AIS data. The method comprises the following steps: preprocessing a historical AIS data set of a ship; dividing the pre-processed AIS data set according to tracks to obtain multiple pieces of trajectory feature data; detecting and deleting error data in the trajectory feature data; performing gridding processing on each piece of trajectory feature data, and dividing prediction targets of trajectory points in each piece of trajectory feature data into eight neighborhood grid directions; and establishing a trajectory prediction model based on an xgboost algorithm, and predicting the trajectory points through a trajectory prediction model. Therefore, the accuracy of ship motion direction prediction is improved. Meanwhile, when the data set is selected, the data can be considered to be valid as long as returned AIS data position information is required to be within a grid precision allowable range, so that the accuracy requirement on the AIS data is reduced.

The invention relates to a reverse pantograph, a charging system and a car. The reverse pantograph comprises a pantograph head, the pantograph head comprises one or more pantograph positive pole plates used for being crossed with positive pole plates of a collector-shoe gear of the car during charging contact, and one or more pantograph negative pole plates used for being crossed with negative pole plates of the collector-shoe gear of the car during charging contact. A positive pole and a negative pole on the pantograph head of the pantograph are correspondingly crossed with and are in buttingcontact with a positive pole and a negative pole of the collector-shoe gear of the car; and in the charging process, the errors in the front, rear, left and right directions of the actual parking position of the car are within the allowable range, the positive pole plates and the negative pole plates of the pantograph can reliably make contact with the positive pole plates and the negative pole plates of the collector-shoe gear correspondingly, thus the precision requirements for the positioning and parking position of the car are lowered, it is ensured that charging is conducted normally, the contact form that the pole plates are arranged in a crossed mode is utilized, the current carrying area is increased, and high-current and high-power charging is ensured.

The invention discloses a mounting structure and a mounting method of a decorative plate. A keel of the mounting structure of the decorative plate is fixed on a wall body, two sides of the lateral wall II of a decorative plate edge hanging piece are hung on the wall of an opening groove of the keel, the lateral wall I of the decorative plate edge hanging piece is hooked on a groove of the decorative plate, and the length direction of the keel is perpendicular to the length direction of the corresponding groove on the decorative plate. The method comprises the following steps of: (1) paying off; (2) fixing the keel; (3) mounting the decorative plate edge hanging piece on the keel; (4) fixing the decorative plate. The edge of the decorative plate and wall body are fixed through the keel and the decorative plate edge hanging piece, and when the edge hanging piece is mounted, the groove fixed on the decorative plate can be utilized, thereby being convenient to mount and easy to implement.

The invention relates to a classification recognition technology for detecting different types of cells through human blood cells. The technology comprises the following steps that A, initial partition is conducted on an original scatter diagram, and subgroup types are determined; B, Gaussian distribution characteristic parameters of all subgroups are calculated respectively, and the maximum log likelihood function value is calculated; C, iteration calculation is entered, and the probability distribution of cells located in the subgroups is calculated according to the Gaussian distribution function output from last iteration; D, Gaussian distribution characteristic parameters of all subgroups are calculated again according to the probability distribution of the cells, and the maximum log likelihood function value is calculated; E, whether the maximum log likelihood function value is convergent or not is judged, and if not, the third step and the fourth step are repeated to conduct iteration calculation; if yes, iteration is stopped, the Gaussian distribution characteristic parameters of all subgroups and the subgroup types of all the cells are output. According to the classification recognition technology for detecting the different types of the cells through the human blood cells, the requirement of accuracy on partition of an initial boundary of the scatter diagram is low, the complexity of the algorithm is reduced, the adaptive capacity is strong, and the accuracy and the stability of particle classification are improved.