Freely selectable small biological sample library control method and system

Abstract

The present invention provides a freely selectable small biological sample library control method and system, which relates to the field of control technology, including establishing a temperature field distribution model by detecting nitrogen concentration data, gas pressure data and temperature data in the sample library, dividing a plurality of temperature independent control areas, and assigning temperature numbers and storage coordinates to each area, and establishing a coordinate database and a path database. Determine the target area according to the identification information and the temperature distribution map, and select the optimal transfer path. Control the robotic arm to transfer the sample along the optimal path, dynamically adjust the clamping force based on force feedback, and adjust the nitrogen flow rate in real time according to the temperature gradient, and finally record the sample storage position, temperature and sample identification information. The present invention realizes the accurate storage and efficient management of samples, improves the safety and reliability of sample storage, and reduces the operating cost of the sample library. In addition, the storage area of ​​the present application can be increased according to the needs of the municipal bureau, and the transmission area can also be increased accordingly.

Description

Technical Field

[0001] The present invention relates to control technology, and in particular to a control method and system for a freely selectable small biological sample library. Background Art

[0002] Small biobanks play a vital role in life science research, clinical diagnosis, and drug development. In order to ensure the integrity and activity of samples, the storage environment of the sample bank needs to be precisely controlled, especially the temperature and gas composition. Traditional sample bank management usually relies on manual operations, which is inefficient and error-prone, and it is difficult to achieve refined management of each sample. With the increase in the number of samples and the diversification of sample types, the demand for automated and intelligent sample bank management systems is becoming increasingly urgent.

[0003] The existing sample library control methods mainly have the following defects and shortcomings:

[0004] 1. Insufficient temperature control accuracy: Traditional sample libraries usually adopt an overall temperature control method, which is difficult to meet the differentiated temperature requirements of different samples, resulting in suboptimal storage temperatures for some samples, affecting sample quality.

[0005] 2. Low efficiency of sample storage and retrieval: Manual operation or simple automated systems are inefficient in the sample storage and retrieval process, especially when dealing with a large number of samples. This is time-consuming and labor-intensive, and increases the risk of samples being exposed to non-ideal environments.

[0006] 3. Lack of refined sample management: Traditional sample library management systems usually only record basic information about samples and lack detailed records of sample storage location, temperature history and other information, making it difficult to achieve full life cycle management of samples. Summary of the invention

[0007] The embodiments of the present invention provide a freely selectable small biological sample library control method and system, which can solve the problems in the prior art.

[0008] According to a first aspect of the embodiments of the present invention,

[0009] Provides optional small biobank control methods, including:

[0010] Detect nitrogen concentration data, gas pressure data and temperature in the sample library to generate initial environmental parameter data; adjust nitrogen flow rate according to the initial environmental parameter data to prevent frost, and establish a temperature field distribution model in combination with multiple temperature control sub-areas; collect temperature distribution data of each storage area, and generate a temperature distribution map in combination with the temperature field distribution model; divide the storage space into multiple temperature-independent control areas according to the temperature distribution map, and assign a temperature number and storage coordinates to each of the temperature-independent control areas; collect the storage coordinates to establish a coordinate database, and plan the movement path of the robot arm according to the coordinate database to generate a path database;

[0011] Obtain image information of samples to be stored, extract sample size, morphological features and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area;

[0012] According to the temperature number and storage coordinates of the target temperature independent control area, the optimal transfer path is selected from the path database; the robotic arm is controlled to transfer the sample along the optimal transfer path, and the clamping force is dynamically adjusted based on the force feedback data during the transfer process; the temperature gradient on the transfer path is calculated according to the temperature distribution map, and the liquid nitrogen flow rate is adjusted in real time to maintain temperature stability; the sample storage location information is recorded, and the storage location information, temperature data and sample identification information are stored in association.

[0013] According to the initial environmental parameter data, the nitrogen flow rate is adjusted to prevent frost from being generated, and a temperature field distribution model is established in combination with multiple temperature control sub-areas; the temperature distribution data of each storage area is collected, and a temperature distribution map is generated in combination with the temperature field distribution model, including:

[0014] The temperature data is collected by a temperature sensor array, the sampling period of the temperature sensor array is 300 milliseconds, and the nitrogen concentration data and the gas pressure data in the equipment are collected at the same time, and the temperature data, the nitrogen concentration data and the gas pressure data are combined into an initial environmental parameter matrix;

[0015] A nitrogen flow regulation model is established according to the initial environmental parameter matrix combined with a characteristic curve of the electromagnetic flow control valve, a PID control algorithm is used to regulate the nitrogen flow, and the opening of the electromagnetic valve is controlled by the nitrogen flow regulation model;

[0016] Divide the storage space into a plurality of temperature control sub-areas, configure each of the temperature control sub-areas with an independent temperature detection unit and a nitrogen injection device, calculate the temperature gradient between the temperature control sub-areas, and establish a temperature field distribution model;

[0017] A temperature monitoring network is constructed by arranging a plurality of temperature sampling points in each of the temperature control sub-areas, and a bicubic spline interpolation algorithm is used to reconstruct a continuous temperature distribution surface according to the temperature data of the temperature sampling points to generate a three-dimensional temperature distribution map including spatial coordinate information, real-time temperature values ​​and temperature change trends.

[0018] Planning the movement path of the robot arm according to the coordinate database, generating the path database includes:

[0019] A path search tree is constructed by adopting an improved RRT algorithm with dynamic bias. The improved RRT algorithm divides the storage space into a target-oriented area and a global search area according to the coordinate database. A polar coordinate system is used to define a sector sampling area in the target-oriented area. The opening angle of the sector sampling area is 60 degrees. The sampling density coefficient of the sector sampling area is set to 0.8. The sampling density of the target-oriented area increases exponentially as the target distance decreases, and the base of the exponential growth is 2. The global search area is divided into uniform grids with a grid side length of 10 cm. Based on the division results of the sampling area, an adaptive step adjustment mechanism is adopted in the node expansion process. The adaptive step range is 2 cm to 8 cm. The adaptive step is linearly positively correlated with the obstacle distance, and the linear coefficient is 0.5 to generate an initial search path.

[0020] For the initial search path, a dynamic model is established through the DH parameters of the joint connecting rod of the robot arm, wherein the DH parameters include the connecting rod angle, the connecting rod length, the connecting rod offset and the joint angle; using the DH parameters, a Newton-Euler recursive algorithm is used to calculate the moment of inertia and the Coriolis force of each joint, the recursive algorithm calculates the joint forces step by step from the end of the robot arm to the base, and establishes a mapping matrix of joint angular acceleration and torque; based on the mapping matrix, a fuzzy controller is used to constrain the movement of the robot arm, the input variables of the fuzzy controller are the joint angular velocity and angular acceleration, the output variable is the angular velocity correction, and the fuzzy rules are used to limit the change rate of the joint angular velocity to within 40 degrees per second, so as to generate a motion path that meets the dynamic constraints;

[0021] Based on the motion path that satisfies the dynamic constraints and combined with the temperature distribution map, a comprehensive path cost function is constructed. The comprehensive path cost function consists of a temperature gradient term, a path length term, and a path curvature term. The weight ratio of each term is 5:3:2. A dynamic programming algorithm is used to minimize the comprehensive path cost function. The state transfer step of dynamic programming is 5 mm, and a path database that satisfies the temperature constraints is obtained.

[0022] The dynamic model is established by the DH parameters of the robot joint link, wherein the DH parameters include the link angle, link length, link offset and joint angle; the Newton-Euler recursive algorithm is used to calculate the moment of inertia and Coriolis force of each joint, and the recursive algorithm calculates the joint force step by step from the end of the robot to the base, and the mapping matrix of the joint angular acceleration and torque is established, including:

[0023] Based on the physical structure of the six-axis robot, a local coordinate system is established for each joint, and four DH parameters, namely, the connecting rod angle, connecting rod offset, connecting rod length and torsion angle, between adjacent joints of the six-axis robot are obtained; the four DH parameters are substituted into a homogeneous transformation matrix, the homogeneous transformation matrix includes rotation transformation terms and translation transformation terms, and the coordinate transformation relationship from the base coordinate system to the end effector coordinate system is obtained by multiplying the homogeneous transformation matrix;

[0024] A forward recursive calculation is performed using a Newton-Euler recursive algorithm, wherein the forward recursive calculation starts from the base coordinate system, calculates the angular velocity and angular acceleration of each connecting rod based on the rotation matrix and the joint axis vector, and calculates the linear acceleration of the connecting rod center of mass according to the connecting rod center of mass position vector;

[0025] Based on the result of the forward recursive calculation, a reverse recursive calculation is performed, wherein the reverse recursive calculation starts from the end effector and uses the mass parameter and inertia tensor of the connecting rod to calculate the force and moment borne by each joint, wherein the calculation of the force and moment includes a gravity term, an inertia term and a Coriolis term;

[0026] The dynamic equation obtained by the reverse recursive calculation is converted into a matrix form, wherein the matrix form includes an inertia matrix, a Coriolis force and a centrifugal force matrix, and a gravity term matrix; the values ​​of the elements of each column of the inertia matrix are calculated by stimulating the acceleration of each joint individually and keeping other joints stationary; the values ​​of the elements of the Coriolis force and centrifugal force matrix are calculated by analyzing the coupling effect of the joint speed; and the values ​​of the elements of the gravity term matrix are calculated by static equilibrium analysis;

[0027] Based on the inertia matrix, the Coriolis force and centrifugal force matrix and the gravity term matrix, a mapping relationship between joint angular acceleration and torque is established, and the mapping relationship is input into the trajectory planning controller and force controller of the robot arm.

[0028] According to the temperature number and storage coordinates of the target temperature independent control area, an optimal transport path is selected from the path database; the robot arm is controlled to transport the sample along the optimal transport path, and the clamping force is dynamically adjusted based on the force feedback data during the transport process, including:

[0029] A path with the smallest value of the comprehensive path cost function is selected from the path database as the optimal transport path; a torque sensor is installed at the end of the robotic arm, the sampling frequency of the torque sensor is 1000 Hz, and the force measurement accuracy is 0.1 Newton; a dynamic impedance model of the end effector is established, and the mass matrix, the damping matrix and the stiffness matrix are updated in real time through an online parameter identification method; for the dynamic impedance model, when the sample weight is less than 100 grams, an impedance characteristic of small stiffness and large damping is adopted, and when the sample weight is greater than 100 grams, an impedance characteristic of large stiffness and small damping is adopted;

[0030] A fuzzy rule base including sample weight, acceleration and temperature is established, wherein the fuzzy rule base uses a BP neural network to realize online learning of rule weights, and the learning rate is set to 0.01; a clamping force correction amount is calculated based on the fuzzy rule base, and a reference clamping force and the clamping force correction amount are added to obtain a real-time clamping force; a force feedback signal is collected by the torque sensor to evaluate the clamping state, and when the force feedback signal is less than a set threshold of 0.5 Newton, a force gain response is triggered within 5 milliseconds;

[0031] The robot arm is controlled according to the optimal path set and the real-time clamping force to perform the sample transfer task, and the temperature fluctuation range is kept within plus or minus 0.5 degrees Celsius and the position accuracy is kept within plus or minus 0.2 millimeters during the transfer process.

[0032] Calculating the temperature gradient on the transport path according to the temperature distribution map, adjusting the nitrogen flow in real time to keep the temperature stable; recording the sample storage location information, and storing the storage location information, temperature data and sample identification information in association, including:

[0033] The transport path is discretized into a sequence of control points with uniform intervals, and the temperature change rate between adjacent control points in the control point sequence is calculated using a central difference format to obtain a temperature gradient vector; based on the numerical value of the temperature gradient vector, adaptive meshing is performed to increase the density of the control point sequence in the area where the temperature changes dramatically;

[0034] A multi-channel nitrogen injection device with a nozzle diameter of two millimeters and a maximum flow rate of five liters per minute is installed in the storage space; a corresponding relationship between the cooling capacity and the nitrogen flow rate is established based on a heat transfer model, and a model predictive control method is used to calculate the temperature tracking error term and the flow rate change penalty term, wherein the prediction time domain is ten seconds and the control time domain is two seconds; an optimal nitrogen flow sequence is obtained by solving a quadratic programming problem online, and the optimal nitrogen flow sequence is sent to the multi-channel nitrogen injection device;

[0035] A multi-level data storage structure including storage location information, temperature data and a sample identification information library is established, wherein the coordinate resolution of the storage location information is 0.1 mm; a ternary relationship graph of samples, locations and temperatures is established using a graph database, wherein the nodes of the ternary relationship graph represent information entities, and the edges of the ternary relationship graph represent association relationships between entities.

[0036] A second aspect of an embodiment of the present invention provides a freely selectable small biological sample library control system, including:

[0037] The first unit is used to detect nitrogen concentration data, gas pressure data and temperature data in the sample library to generate initial environmental parameter data; adjust the nitrogen flow rate according to the initial environmental parameter data to prevent frost, and establish a temperature field distribution model in combination with multiple temperature control sub-areas; collect temperature distribution data of each storage area, and generate a temperature distribution map in combination with the temperature field distribution model; divide the storage space into multiple temperature independent control areas according to the temperature distribution map, and assign a temperature number and storage coordinates to each of the temperature independent control areas; collect the storage coordinates to establish a coordinate database, and plan the movement path of the robot arm according to the coordinate database to generate a path database;

[0038] The second unit is used to obtain image information of the sample to be stored, extract sample size, morphological characteristics and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area;

[0039] The third unit is used to select the optimal transfer path from the path database according to the temperature number and storage coordinates of the target temperature independent control area; control the robotic arm to transfer the sample along the optimal transfer path, and dynamically adjust the clamping force based on the force feedback data during the transfer process; calculate the temperature gradient on the transfer path according to the temperature distribution map, and adjust the liquid nitrogen flow rate in real time to keep the temperature stable; record the sample storage location information, and store the storage location information, temperature data and sample identification information in an associated manner.

[0040] A third aspect of the embodiments of the present invention

[0041] An electronic device is provided, comprising:

[0042] processor;

[0043] a memory for storing processor-executable instructions;

[0044] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0045] According to a fourth aspect of the embodiments of the present invention,

[0046] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0047] The beneficial effects of this application are as follows:

[0048] 1. Precise temperature control: By establishing a temperature field distribution model, dividing temperature independent control areas, and adjusting the nitrogen flow in real time, precise control of the temperature in the sample library is achieved, ensuring the stability of the storage environment for different types of biological samples.

[0049] 2. Automation and efficiency: The robot arm is used to automatically transport samples, and the coordinate database and path database are combined for path planning, which improves the efficiency of sample storage and reduces manual operations. At the same time, the dynamic clamping force adjustment based on force feedback ensures the safety and reliability of the sample transportation process.

[0050] 3. Information management: Sample information is extracted through image recognition and stored in association with temperature, location and other data, thus realizing information management of samples and facilitating the search, statistics and analysis of samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a flowchart of a small biological sample library control method that can be freely selected in an embodiment of the present invention;

[0052] Figure 2 It is a schematic diagram of the structure of a small biological sample library control system that can be freely selected in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0055] Figure 1 FIG. 1 is a flow chart of a small biological sample library control method that can be freely selected in an embodiment of the present invention, such as Figure 1 As shown, the method includes:

[0056] S101. Detect nitrogen concentration data, gas pressure data and temperature data in the sample library to generate initial environmental parameter data; adjust the nitrogen flow rate according to the initial environmental parameter data to prevent frost, and establish a temperature field distribution model in combination with multiple temperature control sub-areas; collect temperature distribution data of each storage area, and generate a temperature distribution map in combination with the temperature field distribution model; divide the storage space into multiple temperature independent control areas according to the temperature distribution map, and assign a temperature number and storage coordinates to each of the temperature independent control areas; collect the storage coordinates to establish a coordinate database, and plan the movement path of the robot arm according to the coordinate database to generate a path database;

[0057] S102. Obtain image information of the sample to be stored, extract sample size, morphological characteristics and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area;

[0058] S103. Select an optimal transfer path from the path database according to the temperature number and storage coordinates of the target temperature independent control area; control the robotic arm to transfer the sample along the optimal transfer path, and dynamically adjust the clamping force based on the force feedback data during the transfer process; calculate the temperature gradient on the transfer path according to the temperature distribution map, and adjust the liquid nitrogen flow rate in real time to maintain temperature stability; record the sample storage location information, and store the storage location information, temperature data and sample identification information in an associated manner.

[0059] In an optional embodiment, adjusting the nitrogen flow rate according to the initial environmental parameter data to prevent frost from being generated, establishing a temperature field distribution model in combination with multiple temperature control sub-areas; collecting temperature distribution data of each storage area, and generating a temperature distribution map in combination with the temperature field distribution model includes:

[0060] The temperature data is collected by a temperature sensor array, the sampling period of the temperature sensor array is 300 milliseconds, and the nitrogen concentration data and the gas pressure data in the equipment are collected at the same time, and the temperature data, the nitrogen concentration data and the gas pressure data are combined into an initial environmental parameter matrix;

[0061] A nitrogen flow regulation model is established according to the initial environmental parameter matrix combined with a characteristic curve of the electromagnetic flow control valve, a PID control algorithm is used to regulate the nitrogen flow, and the opening of the electromagnetic valve is controlled by the nitrogen flow regulation model;

[0062] Divide the storage space into a plurality of temperature control sub-areas, configure each of the temperature control sub-areas with an independent temperature detection unit and a nitrogen injection device, calculate the temperature gradient between the temperature control sub-areas, and establish a temperature field distribution model;

[0063] A temperature monitoring network is constructed by arranging a plurality of temperature sampling points in each of the temperature control sub-areas, and a bicubic spline interpolation algorithm is used to reconstruct a continuous temperature distribution surface according to the temperature data of the temperature sampling points to generate a three-dimensional temperature distribution map including spatial coordinate information, real-time temperature values ​​and temperature change trends.

[0064] First, the storage space is pre-treated, including cleaning, disinfection, and necessary sealing measures to ensure the stability of the environment and reduce the impact of external factors on temperature control. For example, ozone disinfection is used for 30 minutes, followed by ventilation for 2 hours to ensure that the residual ozone concentration is lower than the safety standard.

[0065] Next, deploy the temperature sensor array, nitrogen concentration sensor, and nitrogen pressure sensor. The temperature sensor array is evenly distributed in the storage space, and the sampling period is set to 300 milliseconds. The nitrogen concentration sensor and nitrogen pressure sensor are installed near the nitrogen inlet. For example, in a 10 cubic meter space, 100 temperature sensors can be arranged to form a 10x10 array. Assume that the data collected at a certain moment is: temperature 20°C, nitrogen concentration 98%, and nitrogen pressure 1.2 atmospheres. These data will form the initial environmental parameter matrix.

[0066] Then, a nitrogen flow regulation model is established based on the collected initial environmental parameter matrix and the characteristic curve of the electromagnetic flow control valve. The characteristic curve of the electromagnetic flow control valve describes the relationship between the valve opening and the nitrogen flow. The nitrogen flow regulation model adopts the PID control algorithm to dynamically adjust the opening of the electromagnetic valve according to the difference between the target temperature and the actual temperature to achieve precise nitrogen flow control. For example, when the target temperature is set to 18°C, the PID controller calculates the required nitrogen flow based on the difference between the current temperature of 20°C and the target temperature, and determines the corresponding valve opening according to the characteristic curve of the electromagnetic flow control valve, for example, the opening is 30%.

[0067] The storage space is divided into multiple temperature-controlled sub-areas, each of which is equipped with an independent temperature detection unit and nitrogen injection device. For example, a 10-cubic-meter space is divided into 10 1-cubic-meter sub-areas. By calculating the temperature difference between adjacent sub-areas, the temperature gradient can be obtained, thereby establishing a temperature field distribution model. For example, if the temperature of sub-area 1 is 19°C and the temperature of sub-area 2 is 18.5°C, the temperature gradient between the two sub-areas is 0.5°C / meter.

[0068] Nine temperature sampling points are arranged in each temperature control sub-area to form a temperature monitoring network. For example, in a sub-area of ​​1 cubic meter, nine sampling points can be evenly distributed. Using the bicubic spline interpolation algorithm, a continuous temperature distribution surface is reconstructed based on the temperature data of these nine sampling points. A three-dimensional temperature distribution map is generated by combining spatial coordinate information, real-time temperature values, and temperature change trends. For example, based on the temperature data of nine sampling points, a surface describing the temperature distribution of the sub-area can be generated. The coordinates of each point on the surface represent the spatial position, the color represents the temperature value, and the color change trend represents the temperature change trend.

[0069] This application can achieve:

[0070] Precise temperature control: By establishing a nitrogen flow regulation model and a temperature field distribution model, the temperature of the storage space can be precisely controlled, and the temperature fluctuation can be controlled within a very small range, such as within ±0.5°C. Uniform temperature control: Dividing the storage space into multiple temperature control sub-areas and independently controlling each sub-area can effectively avoid local temperatures that are too high or too low, and ensure the temperature uniformity of the entire storage space, such as a temperature difference of less than 1°C. Visual monitoring: The generated three-dimensional temperature distribution map can intuitively display the temperature distribution in the storage space, allowing users to monitor temperature changes in real time, promptly detect and handle abnormal situations, and improve management efficiency.

[0071] In an optional implementation, planning the movement path of the robot arm according to the coordinate database, generating the path database includes:

[0072] A path search tree is constructed by adopting an improved RRT algorithm with dynamic bias. The improved RRT algorithm divides the storage space into a target-oriented area and a global search area according to the coordinate database. A polar coordinate system is used to define a sector sampling area in the target-oriented area. The opening angle of the sector sampling area is 60 degrees. The sampling density coefficient of the sector sampling area is set to 0.8. The sampling density of the target-oriented area increases exponentially as the target distance decreases, and the base of the exponential growth is 2. The global search area is divided into uniform grids with a grid side length of 10 cm. Based on the division results of the sampling area, an adaptive step adjustment mechanism is adopted in the node expansion process. The adaptive step range is 2 cm to 8 cm. The adaptive step is linearly positively correlated with the obstacle distance, and the linear coefficient is 0.5 to generate an initial search path.

[0073] For the initial search path, a dynamic model is established through the DH parameters of the joint connecting rod of the robot arm, wherein the DH parameters include the connecting rod angle, the connecting rod length, the connecting rod offset and the joint angle; using the DH parameters, a Newton-Euler recursive algorithm is used to calculate the moment of inertia and the Coriolis force of each joint, the recursive algorithm calculates the joint forces step by step from the end of the robot arm to the base, and establishes a mapping matrix of joint angular acceleration and torque; based on the mapping matrix, a fuzzy controller is used to constrain the movement of the robot arm, the input variables of the fuzzy controller are the joint angular velocity and angular acceleration, the output variable is the angular velocity correction, and the fuzzy rules are used to limit the change rate of the joint angular velocity to within 40 degrees per second, so as to generate a motion path that meets the dynamic constraints;

[0074] Based on the motion path that satisfies the dynamic constraints and combined with the temperature distribution map, a comprehensive path cost function is constructed. The comprehensive path cost function consists of a temperature gradient term, a path length term, and a path curvature term. The weight ratio of each term is 5:3:2. A dynamic programming algorithm is used to minimize the comprehensive path cost function. The state transfer step of dynamic programming is 5 mm, and a path database that satisfies the temperature constraints is obtained.

[0075] The robot path planning method aims to generate an optimized path database that meets multiple constraints. The core of this method is to combine the improved rapidly expanding random tree (RRT) algorithm, the robot dynamics model and the temperature distribution map to achieve path planning and optimization.

[0076] First, a coordinate database of the target workspace is established, which contains the precise location information of all obstacles and targets in the robot's operating space.

[0077] Then, using the coordinate database, the robot workspace is divided into a target-oriented area and a global search area. In the target-oriented area, a polar coordinate system is used to define a fan-shaped sampling area with an opening angle of 60 degrees. The sampling density coefficient of this area is initially set to 0.8, and it increases exponentially as the distance to the target decreases. The base of the exponential growth is 2, that is, the closer the distance, the denser the sampling points. For example, if the distance from the current point to the target is 10 cm, the sampling density is 0.8 * 2^(10 / a certain feature distance), where the feature distance is a preset value, such as 5 cm. In the global search area, uniform grid division is used, and the grid side length is set to 10 cm.

[0078] Next, the improved RRT search tree is constructed. During the node expansion process, an adaptive step adjustment mechanism is adopted. The step range is set between 2 cm and 8 cm. The adjustment of the adaptive step is linearly positively correlated with the obstacle distance, with a linear coefficient of 0.5. For example, if the distance from the node to the nearest obstacle is 6 cm, the step is set to 2 + 0.5 * 6 = 5 cm. In this way, the initial search path is generated.

[0079] After obtaining the initial path, the dynamic model of the robot arm is established. Using the DH parameters of the robot arm, such as the connecting rod angle, connecting rod length, connecting rod offset and joint angle, the Newton-Euler recursive algorithm is used to calculate the moment of inertia and Coriolis force of each joint step by step from the end of the robot arm to the base, and establish the mapping matrix of joint angular acceleration and torque.

[0080] Based on the mapping matrix, a fuzzy controller is designed to constrain the motion of the robot arm. The input variables of the fuzzy controller are the joint angular velocity and angular acceleration, and the output variable is the angular velocity correction. Through pre-set fuzzy rules, such as "if the angular acceleration is large, reduce the angular velocity", the change rate of the joint angular velocity is limited to 40 degrees per second, and a motion path that meets the dynamic constraints is generated. Assuming that the angular velocity of a joint is 30 degrees per second and the angular acceleration is 50 degrees per second, according to the fuzzy rules, the controller may output an angular velocity correction of -10 degrees per second to control the angular velocity at the next moment to about 20 degrees per second.

[0081] Finally, combined with the temperature distribution map, a comprehensive path cost function is constructed. This function consists of temperature gradient terms, path length terms, and path curvature terms, and the weight ratio of each term is set to 5:3:2. For example, assuming that the temperature gradient of a certain path is 10 degrees Celsius / cm, the length is 20 cm, and the curvature is 0.1 radians / cm, the cost of this path is 5 * 10 + 3 *20 + 2 * 0.1 = 110.2. The dynamic programming algorithm is used to minimize the cost function with a step size of 5 mm, and finally an optimized path database that meets the temperature constraints is obtained.

[0082] This application can achieve: Improve path planning efficiency: By dividing the sampling area and adopting an adaptive step adjustment mechanism, the search space can be effectively reduced and the path planning efficiency can be improved. Enhance path safety: By establishing a dynamic model and adopting fuzzy control, the movement of the robot arm can be effectively constrained, collisions and singularities can be avoided, and the safety of the path can be enhanced. Optimize path quality: By combining the temperature distribution map and constructing a comprehensive path cost function, the path can be effectively optimized, energy consumption and loss can be reduced, and the path quality can be improved.

[0083] In an optional embodiment, a dynamic model is established by using DH parameters of the robot arm joint link, wherein the DH parameters include link angle, link length, link offset and joint angle; using the DH parameters, a Newton-Euler recursive algorithm is used to calculate the moment of inertia and Coriolis force of each joint, and the recursive algorithm calculates the joint force step by step from the end of the robot arm to the base, and establishing a mapping matrix of joint angular acceleration and torque includes:

[0084] Based on the physical structure of the six-axis robot, a local coordinate system is established for each joint, and four DH parameters, namely, the connecting rod angle, connecting rod offset, connecting rod length and torsion angle, between adjacent joints of the six-axis robot are obtained; the four DH parameters are substituted into a homogeneous transformation matrix, the homogeneous transformation matrix includes rotation transformation terms and translation transformation terms, and the coordinate transformation relationship from the base coordinate system to the end effector coordinate system is obtained by multiplying the homogeneous transformation matrix;

[0085] A forward recursive calculation is performed using a Newton-Euler recursive algorithm, wherein the forward recursive calculation starts from the base coordinate system, calculates the angular velocity and angular acceleration of each connecting rod based on the rotation matrix and the joint axis vector, and calculates the linear acceleration of the connecting rod center of mass according to the connecting rod center of mass position vector;

[0086] Based on the result of the forward recursive calculation, a reverse recursive calculation is performed, wherein the reverse recursive calculation starts from the end effector and uses the mass parameter and inertia tensor of the connecting rod to calculate the force and moment borne by each joint, wherein the calculation of the force and moment includes a gravity term, an inertia term and a Coriolis term;

[0087] The dynamic equation obtained by the reverse recursive calculation is converted into a matrix form, wherein the matrix form includes an inertia matrix, a Coriolis force and a centrifugal force matrix, and a gravity term matrix; the values ​​of the elements of each column of the inertia matrix are calculated by stimulating the acceleration of each joint individually and keeping other joints stationary; the values ​​of the elements of the Coriolis force and centrifugal force matrix are calculated by analyzing the coupling effect of the joint speed; and the values ​​of the elements of the gravity term matrix are calculated by static equilibrium analysis;

[0088] Based on the inertia matrix, the Coriolis force and centrifugal force matrix and the gravity term matrix, a mapping relationship between joint angular acceleration and torque is established, and the mapping relationship is input into the trajectory planning controller and force controller of the robot arm.

[0089] In order to achieve precise control of the six-axis robot, it is necessary to establish its dynamic model, which describes the relationship between joint torque and joint angular acceleration. This method establishes a dynamic model through Denavit-Hartenberg (DH) parameters and Newton-Euler recursive algorithm, and applies the model to trajectory planning and force control.

[0090] First, the six-axis robot arm is modeled. A local coordinate system is established for each joint, and the four DH parameters between adjacent joints are determined: link angle, link offset, link length, and joint angle. For example, the link angle between the first joint and the second joint is 0 degrees, the link offset is 0.1 meters, the link length is 0.5 meters, and the joint angle is variable θ1. Similarly, the DH parameters of all adjacent joints are determined.

[0091] Next, the DH parameters are used to construct a homogeneous transformation matrix. Each homogeneous transformation matrix contains rotational transformation terms and translational transformation terms, which are used to describe the spatial transformation relationship between adjacent coordinate systems. For example, the homogeneous transformation matrix between the first joint and the second joint contains the cosine and sine functions of θ1, as well as translations of 0.1 meters and 0.5 meters. Multiplying the homogeneous transformation matrices of all adjacent joints yields the overall transformation relationship from the base coordinate system to the end effector coordinate system.

[0092] Then, the Newton-Euler recursive algorithm is used for forward recursive calculation. Starting from the base coordinate system, the angular velocity and angular acceleration of each link are calculated step by step. For example, the angular velocity of the first link is equal to the angular velocity of the first joint, and the angular acceleration is equal to the angular acceleration of the first joint. The angular velocity of the second link is equal to the angular velocity of the first link plus the projection of the angular velocity of the second joint in the corresponding coordinate system, and the calculation of angular acceleration is similar. According to the position vector, angular velocity and angular acceleration of the center of mass of each link, the linear acceleration of the center of mass of the link is calculated.

[0093] After completing the forward recursion, perform the reverse recursion calculation. Starting from the end effector, calculate the force and torque of each joint step by step. Gravity, inertia and Coriolis force are considered in the calculation process. For example, the force borne by the sixth joint is equal to the mass of the end effector multiplied by the linear acceleration of the end effector plus gravity. The force borne by the fifth joint is equal to the force transmitted by the sixth joint plus the mass of the fifth connecting rod multiplied by the linear acceleration of the center of mass of the fifth connecting rod plus gravity, and so on. The calculation of torque is similar, and the inertia tensor and Coriolis force of the connecting rod need to be considered.

[0094] The dynamic equation obtained by reverse recursion is converted into matrix form to obtain expressions including inertia matrix, Coriolis force and centrifugal force matrix and gravity matrix. In order to calculate the inertia matrix, the acceleration of each joint can be stimulated separately, the other joints can be kept stationary, and the corresponding joint torque can be measured to determine each element of the inertia matrix. For example, the angular acceleration of the first joint is set to 1 while the other joints are kept stationary. The torque of the first joint is measured, and the torque is the first element of the first column of the inertia matrix. The element values ​​of the Coriolis force and centrifugal force matrices can be calculated by analyzing the coupling effect of the joint velocity. The element values ​​of the gravity matrix can be calculated by static equilibrium analysis. For example, when the robot arm is stationary, the torque of each joint is measured, and these torques are the elements of the gravity matrix.

[0095] Finally, based on the obtained inertia matrix, Coriolis force and centrifugal force matrix, and gravity term matrix, the mapping relationship between joint angular acceleration and torque is established. This mapping relationship can be input into the trajectory planning controller and force controller of the robot arm to achieve precise control of the robot arm.

[0096] This application can achieve:

[0097] Improve control accuracy: Through accurate dynamic models, the required joint torque can be calculated more accurately, thereby improving the trajectory tracking accuracy and force control accuracy of the robot. Enhance stability: The controller based on the dynamic model can better compensate for the nonlinear effects and coupling effects of the robot and improve the stability of the system. Optimize motion performance: Through the dynamic model, the motion trajectory of the robot can be optimized, energy consumption can be reduced, and motion efficiency can be improved.

[0098] In an optional embodiment, according to the temperature number and storage coordinates of the target temperature independent control area, selecting the optimal transport path from the path database; controlling the robot arm to transport the sample along the optimal transport path, and dynamically adjusting the clamping force based on the force feedback data during the transport process includes:

[0099] A path with the smallest value of the comprehensive path cost function is selected from the path database as the optimal transport path; a torque sensor is installed at the end of the robotic arm, the sampling frequency of the torque sensor is 1000 Hz, and the force measurement accuracy is 0.1 Newton; a dynamic impedance model of the end effector is established, and the mass matrix, the damping matrix and the stiffness matrix are updated in real time through an online parameter identification method; for the dynamic impedance model, when the sample weight is less than 100 grams, an impedance characteristic of small stiffness and large damping is adopted, and when the sample weight is greater than 100 grams, an impedance characteristic of large stiffness and small damping is adopted;

[0100] A fuzzy rule base including sample weight, acceleration and temperature is established, wherein the fuzzy rule base uses a BP neural network to realize online learning of rule weights, and the learning rate is set to 0.01; a clamping force correction amount is calculated based on the fuzzy rule base, and a reference clamping force and the clamping force correction amount are added to obtain a real-time clamping force; a force feedback signal is collected by the torque sensor to evaluate the clamping state, and when the force feedback signal is less than a set threshold of 0.5 Newton, a force gain response is triggered within 5 milliseconds;

[0101] The robot arm is controlled according to the optimal path set and the real-time clamping force to perform the sample transfer task, and the temperature fluctuation range is kept within plus or minus 0.5 degrees Celsius and the position accuracy is kept within plus or minus 0.2 millimeters during the transfer process.

[0102] The sample transport method aims to achieve accurate and safe transport of temperature-sensitive samples. First, the system needs to build a path database. The database pre-stores feasible transport paths between different areas and calculates a comprehensive path cost function value for each path. This cost function comprehensively considers factors such as path length, transport time, energy consumption, and temperature stability of the passing area. The smaller the value, the better the overall performance of the path. For example, the cost function value of path A is 1.5, the cost function value of path B is 2.0, and the cost function value of path C is 1.2, then path C is the optimal path.

[0103] Next, the system retrieves all feasible paths from the database based on the target temperature and sample storage coordinates set by the user, and selects the path with the smallest cost function value as the optimal transfer path. Assuming that the sample needs to be transferred from area A (temperature 25°C) to area B (temperature 4°C), the system retrieves three paths, namely path A, path B, and path C, with cost function values ​​of 1.5, 2.0, and 1.2 respectively. The system selects path C with the smallest cost function value as the optimal transfer path.

[0104] To ensure the safety of the samples during transportation, a high-precision torque sensor is installed at the end of the robotic arm, with a sampling frequency of 1000 Hz and a force measurement accuracy of 0.1 Newton. The sensor collects force feedback data in real time and is used to dynamically adjust the gripping force of the robotic arm.

[0105] In order to achieve precise force control, the system establishes a dynamic impedance model of the end effector of the robot arm. The model contains parameters such as mass matrix, damping matrix and stiffness matrix. The system adopts an online parameter identification method to update these parameters in real time according to the feedback data of the torque sensor to adapt to samples of different weights. For example, when the sample weight is 50 grams, the system adjusts the dynamic impedance model to an impedance characteristic with small stiffness and large damping; when the sample weight is 150 grams, the system adjusts the dynamic impedance model to an impedance characteristic with large stiffness and small damping.

[0106] At the same time, the system establishes a fuzzy rule base containing sample weight, acceleration and temperature. The rule base is used to calculate the correction amount of the clamping force. For example, a rule in the rule base can be: if the sample weight is light, the acceleration is small, and the temperature is low, the clamping force correction amount is small. The system uses BP neural network to realize online learning of rule weights, and the learning rate is set to 0.01 to continuously optimize the fuzzy rule base. Assuming that the initial clamping force is 5 Newtons, the clamping force correction amount calculated according to the fuzzy rule base is 0.2 Newtons, then the real-time clamping force is 5.2 Newtons.

[0107] During the sample transfer process, the system continuously collects force feedback signals through the torque sensor and compares them with the set threshold. The threshold is set to 0.5 Newton. When the force feedback signal is less than 0.5 Newton, it indicates that the clamping force is insufficient. The system will trigger a force gain response within 5 milliseconds to increase the clamping force to prevent the sample from slipping.

[0108] Finally, according to the selected optimal path and the clamping force calculated in real time, the robot arm is controlled to perform the sample transfer task. During the transfer process, the system monitors the temperature and position of the sample in real time to ensure that the temperature fluctuation range is within plus or minus 0.5 degrees Celsius and the position accuracy is within plus or minus 0.2 mm.

[0109] This application can achieve:

[0110] Improve transportation efficiency: By pre-establishing a path database and selecting the path with the smallest cost function value, the transportation time can be effectively shortened and the transportation efficiency can be improved. Enhance transportation safety: Dynamically adjust the clamping force based on force feedback data, and combine the fuzzy rule base and dynamic impedance model to effectively prevent the sample from slipping or being damaged, and enhance transportation safety. Improve transportation accuracy: Real-time monitoring of the sample's temperature and position, and precise control, can ensure that the sample maintains a stable temperature and precise position during transportation, improving transportation accuracy.

[0111] In an optional embodiment, the temperature gradient on the transport path is calculated according to the temperature distribution map, and the nitrogen flow rate is adjusted in real time to keep the temperature stable; the sample storage location information is recorded, and the storage location information, temperature data and sample identification information are associated and stored, including:

[0112] The transport path is discretized into a sequence of control points with uniform intervals, and the temperature change rate between adjacent control points in the control point sequence is calculated using a central difference format to obtain a temperature gradient vector; based on the numerical value of the temperature gradient vector, adaptive meshing is performed to increase the density of the control point sequence in the area where the temperature changes dramatically;

[0113] A multi-channel nitrogen injection device with a nozzle diameter of two millimeters and a maximum flow rate of five liters per minute is installed in the storage space; a corresponding relationship between the cooling capacity and the nitrogen flow rate is established based on a heat transfer model, and a model predictive control method is used to calculate the temperature tracking error term and the flow rate change penalty term, wherein the prediction time domain is ten seconds and the control time domain is two seconds; an optimal nitrogen flow sequence is obtained by solving a quadratic programming problem online, and the optimal nitrogen flow sequence is sent to the multi-channel nitrogen injection device;

[0114] A multi-level data storage structure including storage location information, temperature data and a sample identification information library is established, wherein the coordinate resolution of the storage location information is 0.1 mm; a ternary relationship graph of samples, locations and temperatures is established using a graph database, wherein the nodes of the ternary relationship graph represent information entities, and the edges of the ternary relationship graph represent association relationships between entities.

[0115] First, obtain the sample's transport path and discretize it into a series of equally spaced control points. For example, assuming the transport path is a straight line from point A to point B, with the coordinates of point A at (0,0) and point B at (100,100), then this path can be divided into a control point every 10 mm, resulting in 11 control points with coordinates of (0,0), (10,10), (20,20)...,(100,100).

[0116] Next, measure the temperature of each control point. Assume that the measured temperature data are 25℃, 24℃, 23℃, 22℃, 21℃, 20℃, 19℃, 18℃, 17℃, 16℃, and 15℃ respectively.

[0117] Then, the temperature change rate between adjacent control points is calculated to obtain the temperature gradient vector. For example, the temperature change rate between the first control point and the second control point is (24-25)℃ / 10mm = -0.1℃ / mm. Similarly, the temperature change rate between all adjacent control points can be calculated to form a temperature gradient vector.

[0118] Adaptive meshing is performed based on the numerical value of the temperature gradient vector. If the absolute value of the temperature change rate between two adjacent control points is greater than the preset threshold (for example, 0.2°C / mm), the density of control points is increased between the two control points. For example, assuming that the temperature change rate between the fourth control point and the fifth control point is (21-22)°C / 10mm = -0.1°C / mm, which is less than the threshold; and the temperature change rate between the fifth control point and the sixth control point is (20-21)°C / 10mm = -0.1°C / mm, which is also less than the threshold. Then there is no need to increase the density of control points.

[0119] Install multiple nitrogen injection devices in the storage space. Assume that the storage space is a cube with a length, width and height of 1 meter, in which 10x10x10=1000 nitrogen injection devices with a nozzle diameter of 2 mm and a maximum flow rate of 5 liters / minute are installed, evenly distributed in the storage space.

[0120] The corresponding relationship between cooling capacity and nitrogen flow rate is established based on the heat transfer model. Through experiments or simulations, the relationship between nitrogen flow rate and cooling capacity can be obtained. For example, assuming that the cooling capacity is 10 watts when the nitrogen flow rate is 1 liter / minute; the cooling capacity is 20 watts when the nitrogen flow rate is 2 liters / minute, and so on.

[0121] The model predictive control method is used to calculate the temperature tracking error term and the flow change penalty term. For example, assuming that the temperature at the current time t is 20°C and the target temperature is 18°C, the temperature change in the next 10 seconds is predicted and the temperature tracking error term is calculated; at the same time, the energy consumption caused by the change in nitrogen flow rate is considered and the flow change penalty term is calculated. The prediction time domain is 10 seconds and the control time domain is 2 seconds.

[0122] The optimal nitrogen flow sequence is obtained by solving the quadratic programming problem online. According to the predicted temperature change and the set target temperature, the optimal nitrogen flow rate of each nozzle in the next 2 seconds is calculated to minimize the temperature tracking error term and the flow change penalty term.

[0123] The optimal nitrogen flow sequence is sent to the multi-channel nitrogen injection device. For example, the calculated optimal nitrogen flow data for each nozzle in the next 2 seconds is sent to the corresponding nozzle to control its nitrogen injection amount.

[0124] A multi-level data storage structure is established to store sample location information, temperature data, and sample identification information. For example, a relational database is used to store sample identification information, and a graph database is used to store a ternary relationship graph of sample, location, and temperature. The coordinate resolution of the location information is 0.1 mm. For example, the storage location of a sample is (500.0, 500.0, 500.0).

[0125] A graph database is used to build a ternary relationship graph of samples, locations, and temperatures. For example, a sample node is connected to a location node and a temperature node, indicating that the sample is located at the location and the temperature at the location is a specific value.

[0126] This application can achieve:

[0127] Precise temperature control: Through real-time monitoring of temperature gradients and adaptive grid generation, the temperature of the sample storage environment can be controlled more accurately, temperature fluctuations can be controlled within a smaller range, and the quality of sample preservation can be improved. Energy-saving and efficient: The model predictive control method can optimize the nitrogen flow rate, reduce nitrogen consumption and save energy while ensuring temperature stability. Convenient information management: The multi-level data storage structure and graph database can effectively manage a large amount of sample information, location information and temperature data, facilitate query and traceability, and improve management efficiency.

[0128] Figure 2 FIG. 1 is a schematic diagram of a small biological sample library control system that can be freely selected according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0129] The first unit is used to detect nitrogen concentration data, gas pressure data and temperature data in the sample library to generate initial environmental parameter data; adjust the nitrogen flow rate according to the initial environmental parameter data to prevent frost, and establish a temperature field distribution model in combination with multiple temperature control sub-areas; collect temperature distribution data of each storage area, and generate a temperature distribution map in combination with the temperature field distribution model; divide the storage space into multiple temperature independent control areas according to the temperature distribution map, and assign a temperature number and storage coordinates to each of the temperature independent control areas; collect the storage coordinates to establish a coordinate database, and plan the movement path of the robot arm according to the coordinate database to generate a path database;

[0130] The second unit is used to obtain image information of the sample to be stored, extract sample size, morphological characteristics and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area;

[0131] The third unit is used to select the optimal transfer path from the path database according to the temperature number and storage coordinates of the target temperature independent control area; control the robotic arm to transfer the sample along the optimal transfer path, and dynamically adjust the clamping force based on the force feedback data during the transfer process; calculate the temperature gradient on the transfer path according to the temperature distribution map, and adjust the liquid nitrogen flow rate in real time to keep the temperature stable; record the sample storage location information, and store the storage location information, temperature data and sample identification information in an associated manner.

[0132] According to a third aspect of the embodiments of the present invention,

[0133] An electronic device is provided, comprising:

[0134] processor;

[0135] a memory for storing processor-executable instructions;

[0136] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0137] According to a fourth aspect of the embodiments of the present invention,

[0138] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0139] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A freely selectable small biological sample library control method, characterized in that: include: Detect nitrogen concentration data, gas pressure data and temperature data in the sample library to generate initial environmental parameter data; According to the initial environmental parameter data, the nitrogen flow rate is adjusted to prevent frost from being generated, and a temperature field distribution model is established in combination with multiple temperature control sub-areas; the temperature distribution data of each storage area is collected, and a temperature distribution map is generated in combination with the temperature field distribution model; according to the temperature distribution map, the storage space is divided into multiple temperature-independent control areas, and a temperature number and storage coordinates are assigned to each of the temperature-independent control areas; the storage coordinates are collected to establish a coordinate database, and the movement path of the robot arm is planned according to the coordinate database to generate a path database; Obtain image information of samples to be stored, extract sample size, morphological features and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area; According to the temperature number and storage coordinates of the target temperature independent control area, the optimal transfer path is selected from the path database; the robot arm is controlled to transfer the sample along the optimal transfer path, and the clamping force is dynamically adjusted based on the force feedback data during the transfer process; the temperature gradient on the transfer path is calculated according to the temperature distribution map, and the nitrogen flow rate is adjusted in real time to keep the temperature stable; the sample storage location information is recorded, and the storage location information, temperature data and sample identification information are stored in association; A path search tree is constructed by adopting an improved RRT algorithm with dynamic bias. The improved RRT algorithm divides the storage space into a target-oriented area and a global search area according to the coordinate database. A polar coordinate system is used to define a sector sampling area in the target-oriented area. The opening angle of the sector sampling area is 60 degrees. The sampling density coefficient of the sector sampling area is set to 0.

8. The sampling density of the target-oriented area increases exponentially as the target distance decreases, and the base of the exponential growth is 2. The global search area is divided into uniform grids with a grid side length of 10 cm. Based on the division results of the sampling area, an adaptive step adjustment mechanism is adopted in the node expansion process. The adaptive step range is 2 cm to 8 cm. The adaptive step is linearly positively correlated with the obstacle distance, and the linear coefficient is 0.5 to generate an initial search path. For the initial search path, a dynamic model is established through the DH parameters of the robot arm joint link, wherein the DH parameters include the link angle, link length, link offset and joint angle; using the DH parameters, the Newton-Euler recursive algorithm is used to calculate the moment of inertia and Coriolis force of each joint, and the recursive algorithm calculates the joint force step by step from the end of the robot arm to the base, and establishes a mapping matrix between the joint angular acceleration and the torque; Based on the mapping matrix, a fuzzy controller is used to constrain the motion of the robot arm. The input variables of the fuzzy controller are the joint angular velocity and angular acceleration, and the output variable is the angular velocity correction. The fuzzy rule is used to limit the change rate of the joint angular velocity to within 40 degrees per second, and a motion path that meets the dynamic constraints is generated; Based on the motion path that satisfies the dynamic constraints and combined with the temperature distribution map, a comprehensive path cost function is constructed. The comprehensive path cost function consists of a temperature gradient term, a path length term, and a path curvature term. The weight ratio of each term is 5:3:

2. A dynamic programming algorithm is used to minimize the comprehensive path cost function. The state transfer step of dynamic programming is 5 mm, and a path database that satisfies the temperature constraints is obtained.

2. The method according to claim 1, characterized in that According to the initial environmental parameter data, the nitrogen flow rate is adjusted to prevent frost from being generated, and a temperature field distribution model is established in combination with multiple temperature control sub-areas; the temperature distribution data of each storage area is collected, and a temperature distribution map is generated in combination with the temperature field distribution model, including: The temperature data is collected by a temperature sensor array, the sampling period of the temperature sensor array is 300 milliseconds, and the nitrogen concentration data and the gas pressure data in the equipment are collected at the same time, and the temperature data, the nitrogen concentration data and the gas pressure data are combined into an initial environmental parameter matrix; A nitrogen flow regulation model is established according to the initial environmental parameter matrix combined with a characteristic curve of the electromagnetic flow control valve, a PID control algorithm is used to regulate the nitrogen flow, and the opening of the electromagnetic valve is controlled by the nitrogen flow regulation model; Divide the storage space into a plurality of temperature control sub-areas, configure each of the temperature control sub-areas with an independent temperature detection unit and a nitrogen injection device, calculate the temperature gradient between the temperature control sub-areas, and establish a temperature field distribution model; A temperature monitoring network is constructed by arranging a plurality of temperature sampling points in each of the temperature control sub-areas, and a bicubic spline interpolation algorithm is used to reconstruct a continuous temperature distribution surface according to the temperature data of the temperature sampling points to generate a three-dimensional temperature distribution map including spatial coordinate information, real-time temperature values ​​and temperature change trends.

3. The method according to claim 1, characterized in that The dynamic model is established by the DH parameters of the robot joint link, wherein the DH parameters include the link angle, link length, link offset and joint angle; the Newton-Euler recursive algorithm is used to calculate the moment of inertia and Coriolis force of each joint, and the recursive algorithm calculates the joint force step by step from the end of the robot to the base, and the mapping matrix of the joint angular acceleration and torque is established, including: Based on the physical structure of the six-axis robot, a local coordinate system is established for each joint, and four DH parameters, namely, the connecting rod angle, connecting rod offset, connecting rod length and torsion angle, between adjacent joints of the six-axis robot are obtained; the four DH parameters are substituted into a homogeneous transformation matrix, the homogeneous transformation matrix includes rotation transformation terms and translation transformation terms, and the coordinate transformation relationship from the base coordinate system to the end effector coordinate system is obtained by multiplying the homogeneous transformation matrix; A forward recursive calculation is performed using a Newton-Euler recursive algorithm, wherein the forward recursive calculation starts from the base coordinate system, calculates the angular velocity and angular acceleration of each connecting rod based on the rotation matrix and the joint axis vector, and calculates the linear acceleration of the connecting rod center of mass according to the connecting rod center of mass position vector; Based on the result of the forward recursive calculation, a reverse recursive calculation is performed, wherein the reverse recursive calculation starts from the end effector and uses the mass parameter and inertia tensor of the connecting rod to calculate the force and moment borne by each joint, wherein the calculation of the force and moment includes a gravity term, an inertia term and a Coriolis term; The dynamic equation obtained by the reverse recursive calculation is converted into a matrix form, wherein the matrix form includes an inertia matrix, a Coriolis force and a centrifugal force matrix, and a gravity term matrix; the values ​​of the elements of each column of the inertia matrix are calculated by stimulating the acceleration of each joint individually and keeping other joints stationary; the values ​​of the elements of the Coriolis force and centrifugal force matrix are calculated by analyzing the coupling effect of the joint speed; and the values ​​of the elements of the gravity term matrix are calculated by static equilibrium analysis; Based on the inertia matrix, the Coriolis force and centrifugal force matrix and the gravity term matrix, a mapping relationship between joint angular acceleration and torque is established, and the mapping relationship is input into the trajectory planning controller and force controller of the robot arm.

4. The method according to claim 1, characterized in that: According to the temperature number and storage coordinates of the target temperature independent control area, an optimal transport path is selected from the path database; the robot arm is controlled to transport the sample along the optimal transport path, and the clamping force is dynamically adjusted based on the force feedback data during the transport process, including: A path with the smallest value of the comprehensive path cost function is selected from the path database as the optimal transport path; a torque sensor is installed at the end of the robotic arm, the sampling frequency of the torque sensor is 1000 Hz, and the force measurement accuracy is 0.1 Newton; a dynamic impedance model of the end effector is established, and the mass matrix, the damping matrix and the stiffness matrix are updated in real time through an online parameter identification method; for the dynamic impedance model, when the sample weight is less than 100 grams, an impedance characteristic of small stiffness and large damping is adopted, and when the sample weight is greater than 100 grams, an impedance characteristic of large stiffness and small damping is adopted; A fuzzy rule base including sample weight, acceleration and temperature is established, wherein the fuzzy rule base uses a BP neural network to realize online learning of rule weights, and the learning rate is set to 0.01; a clamping force correction amount is calculated based on the fuzzy rule base, and a reference clamping force and the clamping force correction amount are added to obtain a real-time clamping force; a force feedback signal is collected by the torque sensor to evaluate the clamping state, and when the force feedback signal is less than a set threshold of 0.5 Newton, a force gain response is triggered within 5 milliseconds; The sample transfer task is performed according to the optimal transfer path set and the real-time clamping force control robot arm, and the temperature fluctuation range is maintained within plus or minus 0.5 degrees Celsius and the position accuracy is maintained within plus or minus 0.2 millimeters during the transfer process.

5. The method according to claim 1, characterized in that Calculating the temperature gradient on the transport path according to the temperature distribution map, adjusting the nitrogen flow in real time to keep the temperature stable; recording the sample storage location information, and storing the storage location information, temperature data and sample identification information in association, including: The transport path is discretized into a sequence of control points with uniform intervals, and the temperature change rate between adjacent control points in the control point sequence is calculated using a central difference format to obtain a temperature gradient vector; based on the numerical value of the temperature gradient vector, adaptive meshing is performed to increase the density of the control point sequence in the area where the temperature changes dramatically; A multi-channel nitrogen injection device with a nozzle diameter of two millimeters and a maximum flow rate of five liters per minute is installed in the storage space; a corresponding relationship between the cooling capacity and the nitrogen flow rate is established based on a heat transfer model, and a model predictive control method is used to calculate the temperature tracking error term and the flow rate change penalty term, wherein the prediction time domain is ten seconds and the control time domain is two seconds; an optimal nitrogen flow sequence is obtained by solving a quadratic programming problem online, and the optimal nitrogen flow sequence is sent to the multi-channel nitrogen injection device; A multi-level data storage structure including storage location information, temperature data and a sample identification information library is established, wherein the coordinate resolution of the storage location information is 0.1 mm; a ternary relationship graph of samples, locations and temperatures is established using a graph database, wherein the nodes of the ternary relationship graph represent information entities, and the edges of the ternary relationship graph represent association relationships between entities.

6. A freely selectable small biological sample library control system, used to implement the method according to any one of claims 1 to 5, characterized in that: include: The first unit is used to detect nitrogen concentration data, gas pressure data and temperature data in the sample library to generate initial environmental parameter data; According to the initial environmental parameter data, the nitrogen flow rate is adjusted to prevent frost from being generated, and a temperature field distribution model is established in combination with multiple temperature control sub-areas; the temperature distribution data of each storage area is collected, and a temperature distribution map is generated in combination with the temperature field distribution model; according to the temperature distribution map, the storage space is divided into multiple temperature-independent control areas, and a temperature number and storage coordinates are assigned to each of the temperature-independent control areas; the storage coordinates are collected to establish a coordinate database, and the movement path of the robot arm is planned according to the coordinate database to generate a path database; The second unit is used to obtain image information of the sample to be stored, extract sample size, morphological characteristics and sample identification information from the image information; match the sample identification information with the temperature distribution map to determine the target temperature independent control area; The third unit is used to select the optimal transfer path from the path database according to the temperature number and storage coordinates of the target temperature independent control area; control the robotic arm to transfer the sample along the optimal transfer path, and dynamically adjust the clamping force based on force feedback data during the transfer process; calculate the temperature gradient on the transfer path according to the temperature distribution map, and adjust the nitrogen flow rate in real time to keep the temperature stable; record the sample storage location information, and store the storage location information, temperature data and sample identification information in an associated manner.

7. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images