41results about How to "Improve calculation stability" patented technology

The invention relates to a flexible job-shop scheduling optimization method, which applies the Metropolis criterion and the sinusoidal adaptive step length to a firefly algorithm so as to optimize andsolve a discrete problem. On the basis of building a mathematical model, an initial solution population of a discrete combination problem is randomly generated, then local search in an individual domain is performed according to the Metropolis criterion in simulated annealing to generate a new individual, the internal energy difference between the new individual and the original individual is calculated, the new individual is accepted according to a certain probability, and global search is performed on each generation by using the discrete firefly algorithm with the sinusoidal adaptive steplength until an optimal solution is searched. The method can better search an optimal solution of the FJSP (Flexible Job-Shop Scheduling Problem) in the global space and has better search precision, search efficiency and stability, thereby having important significance and significant engineering practical application values for solving discrete problems such as job-shop scheduling.

The invention provides a cloth animation algorithm, which comprises a pretreatment and an animation algorithm. The pretreatment wherein comprises: drive a human-body model by use of data stream in human motion and calculate the current present culmination position of each culmination on a cloth model by using sport data fragments of human; according to the result, present an algorithm for collision detection and response between an active human-body model and a cloth model and record the collision information. In addition, the pretreatment also offers region division for the cloth model which consists of cling region and loose region and binds the cling region on the human-body joints to get a binding value between the culminations of clothes and the joints. The animation algorithm comprises: drive human-body model by use of motion data frames; in the course of moving, adopt a geometrical algorithm to calculate the position of clothes culminations in cling region, and adopt a physical algorithm to calculate the position of clothes culminations in loose region. The invention has the advantages of both high efficiency in geometrical algorithm and verisimilitude in physical algorithm, by which a vivid cloth animation is formed rapidly.

The invention discloses an ellipsoidal boundary Kalman filtering instant positioning and map building method and system, provides a model containing characteristics of set membership filtering and Kalman filtering, and considers model randomness and bounded uncertain factors at the same time. The model takes bounded but uncertain noise into account in a state transition equation and an observationequation, and has better uncertain metric characteristics. In addition, a recursive framework with random uncertainty in the Kalman filter is reserved, and therefore the advantages of the KF are reserved in the iteration process. In a set membership filtering framework, linearization should best adapt to a nonlinear function on a state estimation set rather than only at a state estimation point,and a linearization model is obtained by minimizing a weighted square error between a function value and a linear function approximation on the whole state estimation set. The nonlinear system state variable parameter optimal estimation precision and the system calculation stability are improved through the nonlinear system state variable parameter optimal estimation method.

The present invention discloses a method for detecting the signal in the multi-antenna digital radio communication system; after the emitting terminal in the MIMO system codes the emitted symbols, the symbols reach the receiving terminal through at least one emitting antenna and two different channels; the method for the receiving terminal to detect the symbols emitted by the emitting terminal comprises that the receiving terminal obtains at least two receiving signals and estimates the channel to obtain the channel matrix; and then with the channel matrix, the factorized matrix of the estimated error and covariance of the symbols emitted by the transmitting terminal in the pattern of recursion is calculated, and then based on the obtained factorized matrix, the symbols are detected. With the method disclosed by the present invention, the calculation complexity for detecting signal can be reduced.

The invention discloses a method for determining the position of an erosion line of a hearth of a blast furnace and belongs to the technical field of prolonging service life of the blast furnace. Compared with the prior art, the method for determining the erosion line of the hearth of the blast furnace has the advantages that the calculation amount is small, and accuracy is high. According to the method, a temperature field in a brick line of the hearth is directly calculated based on a steady-state heat conduction equation by a method for stabilizing a grid by using finite elements or finite difference; the heat conductivity coefficient of the brick lining is determined according to the temperature of grid nodes in iterative computation; if the temperature is more than 1,450 DEG C, the temperature field in the brick lining of the hearth is calculated and determined by using the effective heat conductivity coefficient of molten iron; the position of the position of the 1,150-DEG C erosion line is determined according to the temperature field; according to the method for determining the position of the erosion line, the temperature field is only needed to be calculated once; and compared with the conventional method, the method has the advantages that the calculation amount is greatly reduced, the calculation cycle is shortened, various requirements of operation of the blast furnace can be met, the calculation stability is greatly improved, and the division temperature of the molten iron and the brick line is set to be 1,450 DEG C, so that the actual requirement of the blast furnace can be met.

The invention relates to a method for detecting hydrophobic nature of a composite insulator based on dynamic contact angles. The method is characterized by comprising steps of photographing a contact angle meter at the position perpendicular to a plane where an umbrella skirt or a protecting bush is located through a digital camera or a camera lens, an industrial camera and an image acquisition card to obtain a water drop image, continuously shooting through injection of deionized water to obtain a series of water drop images, calculating out an advance angle theta A through a survey meter, then continuously shooting through extraction of water drops to obtain a series of water drop images, and calculating out an advance angle theta R through the survey meter. The method for detecting the hydrophobic nature of the composite insulator based on the dynamic contact angles has the advantages of (1) high calculating accuracy which represents that influences of roughness of the surface of a material and solute settling on the surface can be restrained to some extent, and an receding angle can better reflect the hydrophobic nature of the surface of the material particularly; and (2) high calculating stability which represents that fluctuations among a plurality of measurement results are small, and calculating stability is high due to the fact that the influence of hysteresis phenomena on measurement of the contact angles are restrained.

A displacement detection device includes an image sensor, a light source and a processing unit. The image sensor is configured to successively capture images. The light source provides light with an emission frequency and an emission duration for the image sensor in capturing the images. The processing unit is configured to calculate a displacement according to the images and to adjust both the emission frequency and the emission duration according to the displacement.

The invention discloses a method for elastic scalability configuration in a cloud computing environment. The method comprises the following steps: acquiring current load information of at least one performance indicator dimension in the cloud computing environment; determining whether the load information exceeds a pre-configured load threshold interval, and if so, determining a corresponding loadcurve by using a load information set within the specific time interval in the cloud computing environment, and calculating load change information of the load information contained in the load information set in the specific time interval; and judging whether the load change information is within a pre-configured load change threshold range, and if so, performing elastic scalability configuration on cloud computing resources in the cloud computing environment on the performance indicator dimension. By adoption of the method for elastic scalability configuration in the cloud computing environment, the load information with changes excluding the preset change range in the cloud computing environment can be filtered to save the computing resources or improve the computational efficiency andimproving the computational stability.

The invention proposes an extended holosymmetric multi-cell set membership Kalman hybrid filtering method. The optimal filtering calculation of a system state variable is performed for a nonlinear system model, and the Taylor series polynomial expansion is executed for approaching to a nonlinear system function, and a system linear equivalent model is obtained. The method focuses on the high-ordertruncation error processing of linear operation, achieves the holosymmetric multi-cell approaching calculation of Taylor series high-order errors, and performs the holosymmetric multi-cell set membership filtering. The conventional Kalman filtering is also employed for the Gaussian noise of the system state variable, and the method achieves the holosymmetric multi-cell and Kalman hybrid filteringcalculation. The method improves the optimal estimation precision of a state variable parameter of a nonlinear system and the system calculation stability. Compared with a conventional extended Kalman filtering algorithm, the method is better in calculation advantages and calculation efficiency through an SLAM system simulation experiment.

The invention provides a novel maximum coentropy ellipsoid set membership filtering method, which is used for improving the calculation stability and calculation precision of state parameters of a state space model of a robot system. The invention discloses an ellipsoid set membership filtering algorithm based on a minimum mean square error criterion. A nonlinear system state space model is oriented. The method comprises steps: introducing a novel maximum coentropy MCC criterion in an observation updating step; obtaining a new system noise expression by jointly implementing system model extension operation through a nonlinear system prediction noise error and observation noise; and constructing an error cost function expressed by a second-order information potential energy formula according to a maximum coentropy criterion between a prediction vector and an observation vector based on a system state variable, thereby designing an observation updating calculation process of an ellipsoidset membership filtering algorithm. When the method is used for carrying out land-based robot pose calculation simulation verification, the calculation precision is improved, and the calculation stability is obviously improved compared with a traditional ellipsoid set membership filtering algorithm.

The invention relates to a triangular array-based two-dimensional electronic stability platform real-time calculation method. According to the method, virtual array element simplified calculation is introduced based on an area array triangular grid layout array, so that basic wave control codes can be obtained; ship attitude information and servo information sent by a servo module are received inreal time through a customized optical fiber; a pipeline processing mode is adopted to perform real-time coordinate system correction on a heading, pitch, rolling, servo orientation and triangular inclination angle five-stage pipeline; calculation in the aspects of sine and cosine calculation, wave control code coefficient storage and matrix multiplication is realized through the IP CORE of the Xilinx ISE development software; the significance bit of a wave control code calculation result is automatically discriminated, and therefore, high-precision bit truncation can be realized; after the periodic synchronization signals of a transmission sub-system are received through the optical fiber, a beam scheduling table corresponding to a resource scheduling module is read, and beam control codes of pulse repetition periods in the beam scheduling table are calculated according to beam pointing, frequency points, servo information, and platform correction information.

The invention belongs to the field of aeromagnetic survey, and particularly provides a method and system for improving aeromagnetic data geologic body boundary detection capacity, and the method comprises the steps: S1, obtaining an actually measured or calculated full magnetic gradient tensor data matrix M; s2, calculating according to the full magnetic gradient tensor data matrix M in the step S1 to obtain three characteristic values lambda 1, lambda 2 and lambda 3 of the matrix; s3, calculating according to the full-magnetic gradient tensor data matrix M in the step S1 to obtain a total module value A of the matrix; s4, establishing a boundary detection function E, and obtaining the maximum value max (E) of the E; e = lambda1 * lambda2 * lambda3 * A; and S5, delta max (E) is calculated to obtain a new boundary detection equalization filter iLP, and delta is an adjustment coefficient of equalization depth abnormality. According to the scheme, the method has high resolution, high tilt magnetization resistance and high noise interference resistance, boundary information of aeromagnetic data target geologic bodies with different amplitudes can be balanced, and redundant false boundary anomalies can be avoided. The calculation stability is improved, the signal-to-noise ratio is enhanced, and the transverse resolution can also be improved.

The invention discloses an unconditionally stable multi-point excitation collapse explicit analysis method for a super-large-span bridge. The unconditionally stable multi-point excitation collapse explicit analysis method includes the steps: establishing a finite element model of the structure and defining a fiber cross section, then converting the uniformly distributed load of the structure intocentralized quality processing, and obtaining a seismic response image by using a KR algorithm. The unconditional stable explicit integral algorithm applied in the invention can make up the problems of non-convergence caused by iteration, too small step length in a loading path or display integral algorithm, low efficiency precision and the like of the implicit integral method to a great extent. The unconditionally stable multi-point excitation collapse explicit analysis method can obtain a stable solution without time step length requirements, and is high in the efficiency is high when a large-span structure and a complex structure.

The invention discloses a Tilt method aeromagnetic boundary detection method and device based on tensor eigenvalues and a storage medium. The method comprises the steps of acquiring a full magnetic gradient tensor data matrix M; establishing a boundary detection function E and a depth resolution gain factor Mz according to the full magnetic gradient tensor data matrix M; respectively calculating the horizontal gradient Ex of E in the x direction, the horizontal gradient Ey of E in the y direction and the vertical gradient Ez of E in the z direction; calculating a total gradient mode ASM according to the x-direction horizontal gradient Ex, the y-direction horizontal gradient Ey and the z-direction vertical gradient Ez; and according to the z-direction vertical gradient total gradient mode ASM, the depth resolution gain factor Mz and the full magnetic gradient tensor data matrix M, implementing aeromagnetic boundary detection. According to the Tilt method aeromagnetic boundary detection method and device based on tensor eigenvalues and the storage medium provided by the invention, the problem of singular point analysis existing in the Tilt method can be solved, the calculation stability is improved, false interference information is eliminated, the signal-to-noise ratio is enhanced, and the aeromagnetic data processing conversion quality and the geologic body boundary identification effect are improved.

The invention relates to the field of crop canopy image recognition, particularly to a rice canopy recognition method based on a digital camera image. The method comprises: (1), 1) obtaining a by using a digital camera; (2), obtaining color values of R, G, and B channels in the color rice canopy image; and (3), determining whether the image is a rice canopy according to the R, G, and B color values in each pixel of the color rice canopy image. Therefore, the identification rate of the method is high; a working step of selecting an image sample in advance according to the existing common method can be removed; the calculation process is simple; and the work efficiency can be improved obviously.

The invention discloses a method and device for obtaining accurate aeromagnetic survey data based on a Tilt method, and the method comprises the steps: calculating the horizontal gradient of a horizontal gradient in an x direction in a y direction and the vertical gradient in a z direction according to the aeromagnetic survey data T, and calculating the total horizontal gradient THDR according to the horizontal gradient in the x direction and the horizontal gradient in the y direction, and calculating a z-direction vertical gradient average value Wmean (VDR) and a total horizontal gradient average value Wmean (THDR) according to the z-direction vertical gradient and the total horizontal gradient THDR, and obtaining accurate aerial magnetic measurement data according to the z-direction vertical gradient average value VDR, the total horizontal gradient THDR, the z-direction vertical gradient average value Wmean (VDR) and the total horizontal gradient average value Wmean (THDR). According to the method and device for obtaining the accurate aeromagnetic survey data and the storage medium, the problem of singular point analysis existing in a Tilt method can be solved, the calculation stability is improved, boundary interference of geologic bodies generating false aeromagnetic data is eliminated, the signal-to-noise ratio is increased, and the aeromagnetic data processing and conversion quality and the geologic body boundary recognition effect are improved.

The invention discloses an axial flux-based quasi-three-dimensional transport calculation method and system, and the method comprises the steps: building a three-dimensional neutron transport equation, and enabling an input item of the three-dimensional neutron transport equation to be advanced reactor data; and in the axial direction, carrying out flux expansion in a first-order difference mode and zero setting operation on the appearing negative flux, calculating axial and radial leakage items of each flat source region according to the flux of the boundary surface after zero setting, and calculating the standard flux of each flat source region according to the axial and radial leakage items. The negative flux is zeroed, a boundary flow calculation leakage item is obtained through statistics after the negative flux is zeroed, and the standard flux of the flat source area is recalculated through the leakage item. Therefore, the situation that flow and flux are not matched due to simple zero setting, and calculation convergence is affected is avoided.

The invention discloses a parameterized mesh generation method for a unidirectional composite material mesoscopic finite element model. The method is characterized by comprising the following steps of: establishing the unidirectional composite material mesoscopic geometric model comprising a matrix, an interface layer and fiber monofilaments; dividing the mesoscopic geometric model into an external matrix area and an internal square block area; segmenting the internal square block body area; and defining a division parameter represented by the number of segments as a reference of the mesh division density of the model so as to control the overall mesh density and the like of the model. According to the method, the divided grids are regular in shape and approximate in size, so that the accuracy of a finite element calculation result can be ensured; orderly change of grids in each region is guaranteed through introduced division constraints, the situation of grid change imbalance is avoided, and the calculation stability can be improved; meanwhile, the unified control of the overall grid density of the model is realized by taking a shared division parameter as a reference, and the adjustment efficiency of the grid density of the model is high.

The invention discloses a real-time detection method of a digital image correlation method based on SIFT improvement, and relates to a deformation detection method. The method comprises the following steps: 1, obtaining gray level images before and after deformation; 2, calculating sub-pixel curved surface fitting; 3, calculating a sub-pixel interpolation; 4, establishing different Gaussian scale spaces to construct a continuous scale image pyramid; 5, searching and screening feature points from the continuous scale space; 6, constructing analysis sub-regions in combination with the selected feature points; 7, calculating the displacement of the deformation sub-region; and 8, using a displacement feature vector to represent deformation. According to the method, the continuity of the displacement field is ensured, the adjustment of the calculation speed and precision is more concise and convenient, the two parameters are balanced, the improved algorithm does not increase the calculation complexity, the pretreatment time is saved, the method has better adaptability to complex boundaries, and higher calculation precision can be achieved under the condition of the same calculation speed.

The invention belongs to the technical field of meteorological prediction, and particularly relates to a forecasting system and method for urban rainstorm ponding simulation real-time verification. The invention aims to solve the technical problem of a forecasting mode of urban rainstorm real-time certification, and provides a forecasting method of urban rainstorm real-time certification. Model calculation and accumulated water monitoring assimilation processing are adopted to determine a local production and confluence coefficient and a drainage coefficient, and the model is a mathematical mode which is completed through assimilation parameter verification based on data assimilation of a model calculation result and an accumulated water monitoring result; the provided mode can improve the precision of urban rainstorm forecast simulation, the calculation efficiency can be improved by correcting the production and confluence coefficient and the drainage coefficient in real time, and the calculation stability can be improved by adopting model calculation and accumulated water monitoring assimilation processing.

The invention discloses an ellipsoidal boundary Kalman filtering instant positioning and map building method and system, provides a model containing characteristics of set membership filtering and Kalman filtering, and considers model randomness and bounded uncertain factors at the same time. The model takes bounded but uncertain noise into account in a state transition equation and an observationequation, and has better uncertain metric characteristics. In addition, a recursive framework with random uncertainty in the Kalman filter is reserved, and therefore the advantages of the KF are reserved in the iteration process. In a set membership filtering framework, linearization should best adapt to a nonlinear function on a state estimation set rather than only at a state estimation point,and a linearization model is obtained by minimizing a weighted square error between a function value and a linear function approximation on the whole state estimation set. The nonlinear system state variable parameter optimal estimation precision and the system calculation stability are improved through the nonlinear system state variable parameter optimal estimation method.

The invention discloses an SOC estimation method, a mobile power supply and a readable storage medium, mainly by calculating the power value of the mobile power supply with a fixed step size, and then step by step and superimposing the unit power to calculate the cumulative distance from the mobile power supply to the last SOC change time. Electricity, then compare the absolute value of the accumulated electricity with the electricity value of the fixed step size, and based on the comparison result, add / decrease the SOC value by a fixed step size or keep the SOC value unchanged, and obtain the starting value of the accumulated electric quantity at the next moment . By obtaining the SOC value of the mobile power supply in the above way, in actual programming, only if judgment and addition and subtraction operations can be used to replace the division operation of the existing SOC estimation method, which greatly reduces the amount of calculation when estimating SOC and reduces the consumption of computing resources. As well as the occupation of storage space, it greatly reduces the requirements for chip computing power and peripheral resources, reduces equipment energy consumption, improves SOC estimation speed and improves computing stability.

The invention discloses an aeromagnetic boundary detection method, device and storage medium based on the tensor eigenvalue Theta Map method. The method includes: obtaining a full magnetic gradient tensor data matrix M; establishing a boundary detection function E and a depth resolution gain according to M Factor M z ;Respectively calculate the horizontal gradient E of E in the x direction x , the horizontal gradient E in the y direction y and the vertical gradient E in the z direction z ;according to E x ,E y and E z Calculate the total horizontal gradient THDR and the total gradient modulus ASM; according to THDR, ASM, M z , M, realize aeromagnetic boundary detection. The aeromagnetic boundary detection method, device and storage medium of the Theta Map method based on tensor eigenvalues ​​provided by the present invention can solve the problem of "analytic singularity" in the Theta Map method, improve calculation stability, eliminate false interference information, and enhance the signal-to-noise ratio. And improve the quality of aeromagnetic data processing and transformation and the recognition effect of geological body boundaries.

The invention relates to a wind-resistant design method of a cable-membrane structure based on strong coupling overall technology, which establishes the initial form of the cable-membrane structure; establishes the control equation of the fluid-solid coupling system and the coupling conditions at the fluid-solid interface, and performs time-space discretization on it, Obtain the strong coupling overall equation of the fluid-solid coupling system; use the SST K‑ω turbulence model to simulate turbulence, obtain the strong coupling overall equation of the fluid-solid coupling system considering the turbulence model, and use the Newton method to linearize it; after constructing the linearized The preprocessor of the strong coupling global equation matrix, obtains the strong coupling global equation of the fluid-solid coupling system with the preprocessor, uses the Krylov subspace projection method to solve it, and obtains the strong coupling global equation fluid pressure, fluid velocity, The displacement of the cable-membrane structure and the displacement of the linear elastic finite element model; thus, the wind-resistant design of the cable-membrane structure is carried out, and the wind-resistant design parameters are obtained.

The invention provides a bearing dynamic performance analysis method considering the influence of a rotor and an elastic support, and the method comprises the steps: calculating the acting force of a rolling body and a ferrule through combining the force balance relation of a bearing-rotor-elastic support system; calculating the acting force of the rolling body, the retainer and the lubricating oil in combination with oil supply conditions; solving a rolling body and retainer motion differential equation to obtain rolling body and retainer motion; solving the corrected motion states of the rolling body and the retainer according to a motion differential equation of the supporting substructure, and calculating the performance parameters of the bearing at the moment and the constraint reaction force of the supporting substructure; solving a rotor motion differential equation to obtain rotor displacement and speed response; and substituting the displacement of the rotor node where the bearing is located into the bearing mechanical model, and repeating the process to obtain the time domain response of the bearing-rotor-elastic support system in a preset time period. According to the method, the coupling relation among the lubrication effect, the holder vibration, the rotor inertia and the elastic support deformation is comprehensively considered, and a theoretical method is provided for introducing the influence of the elastic support and the rotor during bearing performance simulation.

The invention discloses an aeromagnetic boundary detection method, device and storage medium of the Tilt method based on tensor eigenvalues. The method includes: acquiring a full magnetic gradient tensor data matrix M; establishing a boundary detection function according to the full magnetic gradient tensor data matrix M E and depth resolution gain factor M z ;Respectively calculate the horizontal gradient E of E in the x direction x , the horizontal gradient E in the y direction y and the vertical gradient E in the z direction z ;According to the horizontal gradient E in the x direction x , the horizontal gradient E in the y direction y and the vertical gradient E in the z direction z Calculate the total gradient modulus ASM; according to the total gradient modulus ASM of the vertical gradient in the z direction, the depth resolution gain factor M z , The full magnetic gradient tensor data matrix M realizes aeromagnetic boundary detection. The aeromagnetic boundary detection method, device and storage medium of the Tilt method based on tensor eigenvalues ​​provided by the present invention can solve the problem of "analytical singularity" in the Tilt method, improve calculation stability, eliminate false interference information, enhance signal-to-noise ratio, and improve The quality of aeromagnetic data processing and conversion and the effect of geological body boundary recognition.

The invention relates to a coupling calculation method and system. The method includes the following steps that a coupling calculation service module generates a plurality of first independent variables, and the first independent variables are transmitted to corresponding independent calculation modules respectively; the independent calculation modules calculate corresponding residual errors, and derivatives, at the first independent variables, of equations to be solved are calculated according to an automatic differentiation method; the coupling calculation service module carries out calculation to obtain a plurality of second independent variables; the independent calculation modules judge whether the second independent variables are solutions of the corresponding equations to be solved or not, if not, circular calculation is carried out until calculated independent variables are the solutions of the corresponding equations to be solved, and if yes, the solutions are output. The derivative, at the corresponding first independent variable, of each equation to be solved is calculated by the corresponding independent calculation module according to the automatic differentiation method, high-precision calculation results can be obtained, the number of follow-up iterations of the coupling calculation service module is reduced to a great degree, and the calculation stability of the coupling calculation service module is greatly improved.

The invention provides a cloth animation algorithm, which comprises a pretreatment and an animation algorithm. The pretreatment wherein comprises: drive a human-body model by use of data stream in human motion and calculate the current present culmination position of each culmination on a cloth model by using sport data fragments of human; according to the result, present an algorithm for collision detection and response between an active human-body model and a cloth model and record the collision information. In addition, the pretreatment also offers region division for the cloth model whichconsists of cling region and loose region and binds the cling region on the human-body joints to get a binding value between the culminations of clothes and the joints. The animation algorithm comprises: drive human-body model by use of motion data frames; in the course of moving, adopt a geometrical algorithm to calculate the position of clothes culminations in cling region, and adopt a physicalalgorithm to calculate the position of clothes culminations in loose region. The invention has the advantages of both high efficiency in geometrical algorithm and verisimilitude in physical algorithm,by which a vivid cloth animation is formed rapidly.

The present invention discloses a method for detecting the signal in the multi-antenna digital radio communication system; after the emitting terminal in the MIMO system codes the emitted symbols, the symbols reach the receiving terminal through at least one emitting antenna and two different channels; the method for the receiving terminal to detect the symbols emitted by the emitting terminal comprises that the receiving terminal obtains at least two receiving signals and estimates the channel to obtain the channel matrix; and then with the channel matrix, the factorized matrix of the estimated error and covariance of the symbols emitted by the transmitting terminal in the pattern of recursion is calculated, and then based on the obtained factorized matrix, the symbols are detected. Withthe method disclosed by the present invention, the calculation complexity for detecting signal can be reduced.

The invention discloses a control system of a trapezoidal speed curve of a 3D printer extrusion head. A brand-new trapezoidal speed curve control mechanism is included. Different from the traditional manner of achieving trapezoidal speed curve control through a firmware program, the control system achieves speed control based hardware. The defect that a timer interrupt routine is needed for driving the extrusion head to operate in the original solution is abandoned, all computational processes are achieved through the hardware IP core (intellectual property core), the system is divided into several function submodules, and the function submodules are responsible for different functions. Through hardware-based design, when the trapezoidal speed curve of the 3D printer extrusion head is output, the computational processes are simplified, the load of a CPU is released, computational stability is enhanced, computational time is reduced, and contributions are made to improving the printing speed of a 3D printer.