Posture detection, synchronous lifting and fine adjustment control system and process for gallery hoisting

Abstract

The present application relates to the field of automation control technology, and discloses a posture detection, synchronous lifting and fine adjustment control system and process for gallery hoisting, comprising: obtaining hoisting point cable parameter and dynamic information to calculate dynamic equivalent stiffness, reconstructing stress centroid and decoupling natural deflection and torsion deformation; then constructing a load surge penalty functional based on stiffness and deformation parameters, obtaining optimal speed command through colony information entropy driven particle swarm iteration; combining real-time wind speed to reshape target command tension, substituting into the fluid control equation to inversely solve the servo valve port mechanical opening area; finally, based on real-time tension and stiffness, calculating the background natural vibration period and evolving the control step size. The present application can resolve the hyperstatic risk caused by Venturi wind field disturbance and suspension cable hardening, and improve the safety of high-altitude operation.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, and more specifically, to a control system and process for attitude detection, synchronous lifting and fine-tuning of connecting corridors during hoisting. Background Technology

[0002] The overall hoisting of long-span connecting corridors between high-rise buildings is an extremely challenging engineering task. Due to the often asymmetrical arrangement of pipelines and equipment within these corridors, their geometric center naturally deviates from their center of mass and torsional rigidity. When operating within the canyon-like space between supertall buildings, the airflow within the passageway is easily compressed, generating a Venturi effect that causes a nonlinear increase in wind speed with height, applying highly uneven aerodynamic torque to the suspended corridor. Existing multi-point synchronous lifting systems typically rely on control logic based on absolute synchronization of rigid body displacement, forcing all hydraulic lifting devices to maintain a consistent displacement rate. However, during lifting operations, as the length of the suspension cables continuously shortens, the overall equivalent stiffness of the hoisting system exhibits a sharp increasing hardening trend. Under this extremely stiff high-altitude condition, the flexible torsional deformation that inevitably occurs in the corridor under complex wind fields cannot be naturally released. At this point, the forced displacement synchronization command leads to a vicious misalignment and intersection between the strain energy generated by wind loads and the forced displacement energy, generating extremely high parasitic strain energy within the system. This causes a nonlinear surge in the load on some support points, resulting in serious safety hazards such as tearing of the corridor structure or cable breakage. Summary of the Invention

[0003] This invention provides a posture detection, synchronous lifting and fine-tuning control system and process for hoisting connecting corridors, which solves the technical problems mentioned in the background art.

[0004] like Figure 1 As shown, the attitude detection, synchronous lifting, and fine-tuning control technology for the hoisting of the connecting corridor is applied to a control system that includes multi-point suspended cables and hydraulic actuators with proportional valves. This process is executed within the current control cycle and includes: The free length of the cable, the absolute displacement vector and the real-time tension vector of each suspension point are obtained, and the dynamic equivalent stiffness of each suspension point is adaptively calculated. The dynamic force centroid is reconstructed using the real-time tension vector and the absolute displacement vector of each of the aforementioned suspension points, and the natural torsional deformation vector is orthogonally decoupled from the absolute displacement vector of each of the aforementioned suspension points. A candidate command cluster containing multiple candidate velocity command vectors is generated. Based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points, a load surge penalty functional for evaluating each of the candidate velocity command vectors is constructed. The community information entropy of the candidate command community is calculated based on the load surge penalty functional, and the optimal speed command vector is calculated through optimization iteration. Real-time wind speed is obtained to calculate the total Venturi wind driving force, and the target command tension is generated by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. Substituting the target command force into the fluid control equation, the mechanical opening area used to control the corresponding proportional valve is obtained by inverse solution; Based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, the natural vibration period is calculated, and the control step size of the next period is obtained by combining the community information entropy.

[0005] Secondly, the attitude detection, synchronous lifting, and fine-tuning control system for the hoisting of the connecting corridor, realizing the attitude detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor as described in any one of the claims, includes: The stiffness calculation module is used to obtain the cable free length, absolute displacement vector and real-time tension vector of each suspension point within the current control cycle, and adaptively calculate the dynamic equivalent stiffness of each suspension point. The deformation decoupling module is used to reconstruct the dynamic force centroid using the real-time tension vector and the absolute displacement vector of each of the suspension points, and to orthogonally decouple the natural torsional deformation vector from the absolute displacement vector of each of the suspension points. A functional construction module is used to generate a candidate command cluster containing multiple candidate velocity command vectors, and to construct a load surge penalty functional for evaluating each candidate velocity command vector based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points. The optimization module is used to calculate the community information entropy of the candidate command community based on the load surge penalty functional, and to calculate the optimal speed command vector through optimization iteration. The tension generation module is used to obtain real-time wind speed to calculate the total Venturi wind driving force, and to generate the target command tension by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. The valve port control module is used to substitute the target command pull force into the fluid control equation and solve it inversely to obtain the mechanical opening area used to control the corresponding proportional valve. The step size evolution module is used to calculate the background natural vibration period based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, and to obtain the control step size of the next period by combining the community information entropy.

[0006] The beneficial effects of this invention include: by solving the dynamic equivalent stiffness of the suspension point in real time and orthogonally decoupling the natural torsional deformation from the displacement, constructing a load surge penalty functional to guide the particle swarm optimization algorithm, and supplementing it with wind field feedforward force reshaping and valve port absolute area mapping, it not only effectively avoids the danger of statically indeterminate load mutation caused by rigid body forced synchronization, but also realizes the precise allocation of the absolute mechanical opening of the hydraulic actuator and the autonomous evolution of the control cycle under complex aerodynamic disturbances, ensuring the system safety and coordinated stability of the multi-point suspension lifting process. Attached Figure Description

[0007] Figure 1 This is a flowchart of the posture detection, synchronous lifting and fine-tuning control process for the hoisting of the connecting corridor according to the present invention. Detailed Implementation

[0008] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0009] Example 1: As Figure 1 As shown, the attitude detection, synchronous lifting, and fine-tuning control technology for the hoisting of the connecting corridor is applied to a control system that includes multi-point suspended cables and hydraulic actuators with proportional valves. This process is executed within the current control cycle and includes: The free length of the cable, the absolute displacement vector and the real-time tension vector of each suspension point are obtained, and the dynamic equivalent stiffness of each suspension point is adaptively calculated. The dynamic force centroid is reconstructed using the real-time tension vector and the absolute displacement vector of each of the aforementioned suspension points, and the natural torsional deformation vector is orthogonally decoupled from the absolute displacement vector of each of the aforementioned suspension points. A candidate command cluster containing multiple candidate velocity command vectors is generated. Based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points, a load surge penalty functional for evaluating each of the candidate velocity command vectors is constructed. The community information entropy of the candidate command community is calculated based on the load surge penalty functional, and the optimal speed command vector is calculated through optimization iteration. Real-time wind speed is obtained to calculate the total Venturi wind driving force, and the target command tension is generated by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. Substituting the target command force into the fluid control equation, the mechanical opening area used to control the corresponding proportional valve is obtained by inverse solution; Based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, the natural vibration period is calculated, and the control step size of the next period is obtained by combining the community information entropy.

[0010] This invention is applied to a control system that includes multi-point suspended cables and hydraulic actuators with proportional valves. The hardware carrier of the control system includes a three-dimensional attitude sensor, anemometer, and GNSS positioning device installed on the main body of the connecting corridor; stroke sensors, pin-type tension sensors, and oil pressure sensors installed on the hydraulic cylinders at each suspension point; a pressure sensor installed on the hydraulic pump station; and an industrial controller and a data acquisition module. The operation cycle of the industrial controller is not less than 100Hz, and the sampling frequency of the data acquisition module is not less than 200Hz. All acquisition devices are connected through an industrial Ethernet bus and clock synchronization is achieved using the IEEE 1588 precision time protocol, with a synchronization error of no more than 1 millisecond.

[0011] Before the hoisting operation begins, the initial parameter calibration process must be completed. With the connecting corridor initially positioned on the ground, the three-dimensional coordinates of each hoisting point in the independent construction coordinate system are measured using a total station, serving as the initial reference coordinates for hoisting. The unit is meters. The construction independent coordinate system uses the preset benchmark point of the hoisting operation area as the origin, with the X-axis along the length of the connecting corridor, the Y-axis along the width of the connecting corridor, and the Z-axis along the vertical upward direction. Measure the external dimensions of the connecting corridor body, and in conjunction with the structural design drawings, calculate the windward projected area of ​​the connecting corridor perpendicular to the wind direction under different wind angles. The data, in square meters, is stored in the industrial controller's storage unit. The specifications of the hoisting cables, including the cable's modulus of elasticity, are retrieved. Cable cross-sectional area Obtain hydraulic system parameters, including the effective area of ​​the hydraulic cylinder. Oil elastic modulus Oil density Flow coefficient of proportional valve All fixed parameters are pre-stored in the industrial controller.

[0012] Within each control cycle, the industrial controller sequentially performs data acquisition, parameter calculation, instruction generation, control output, and step size update operations according to a fixed time sequence. The specific implementation process is as follows.

[0013] The first step is to obtain the free length of the cable, the absolute displacement vector, and the real-time tension vector at each suspension point, and then adaptively calculate the dynamic equivalent stiffness of each suspension point.

[0014] Obtain the cable elastic modulus and cross-sectional area of ​​the multi-point suspended cable, as well as the absolute displacement vector and real-time tension vector of the previous control cycle. Cable elastic modulus This is the ratio of normal stress to corresponding normal strain in the cable material during the elastic deformation stage, expressed in Pascals (Pa), ranging from 1.9 × 10¹¹ Pa to 2.1 × 10¹¹ Pa, determined by the cable's factory inspection report. Cable cross-sectional area. The nominal cross-sectional area of ​​the cable, in square meters, is determined by the cable specifications. The absolute displacement vector and real-time tension vector of the previous control cycle are the effective acquired data, latched and filtered, at the end of the previous control cycle and stored in the register of the industrial controller.

[0015] Obtain the cable free length, absolute displacement vector, and real-time tension vector for each suspension point within the current control cycle. Cable free length Let be the length of the free section of the cable at the i-th lifting point at time t, in meters. This is calculated by converting the cylinder extension stroke data collected in real-time by the hydraulic cylinder's stroke sensor. 'i' is the lifting point index, a positive integer from 1 to n, and 'n' is the total number of lifting points in the connecting corridor, determined by the corridor's structural form and lifting plan. (Absolute displacement vector) Let be the three-dimensional coordinate vector of the i-th lifting point at time t in the construction independent coordinate system, in meters. This vector is synchronously acquired using GNSS positioning equipment installed at the lifting point. The acquired data undergoes moving average filtering, with a filtering window length of 3 to 5 sampling points. Real-time tension vector. The vector formed by the tension value of the i-th suspension point along the cable axis at time t, in Newtons, is obtained by directly acquiring the pressure data through a pin-type tension sensor or by multiplying the pressure data acquired by the oil pressure sensor in the rodless chamber of the hydraulic cylinder by the effective area of ​​the hydraulic cylinder. The two acquisition methods can be used for mutual redundancy verification.

[0016] Divide the free length of the cable at each suspension point by the product of the cable's elastic modulus and cross-sectional area to obtain the cable's tensile flexibility term. The calculation formula is as follows: The unit of cable tensile flexibility is meters per Newton (m / N), which characterizes the deformation flexibility of a cable under theoretical elastic tension and is the basic flexibility component of the cable's inherent properties.

[0017] Calculate the absolute value of the difference between the absolute displacement vector of each lifting point and the absolute displacement vector of the previous control cycle to obtain the displacement difference modulus. The calculation formula is as follows: in, Let be the absolute displacement vector of the i-th lifting point in the previous control cycle. This represents the step size of the current control cycle, in seconds. The modulus operation is the 2-norm of a vector. The unit of the displacement difference modulus is meters, which represents the actual displacement change of the lifting point within a single control cycle.

[0018] Calculate the absolute value of the difference between the real-time tension vector at each lifting point and the real-time tension vector of the previous control cycle to obtain the tension differential modulus. The calculation formula is as follows: in, This is the real-time tension vector of the i-th lifting point in the previous control cycle. The unit of the tension differential modulus is Newtons, which represents the actual change in cable tension at the lifting point within a single control cycle.

[0019] Dividing the displacement differential modulus by the tensile force differential modulus yields the dynamic elastic compliance term, calculated using the following formula: The unit of dynamic flexibility is meters per Newton (m / N), which characterizes the actual deformation flexibility of the cable and hydraulic system under the current working conditions. When the tensile differential modulus is 0, the dynamic flexibility term is set to 0 to avoid the error of dividing by 0. When there is a jump in the collected data, the average value of the displacement differential modulus and the tensile differential modulus of three adjacent control cycles is used for calculation to suppress the influence of data noise.

[0020] The dynamic equivalent stiffness of each suspension point is obtained by adding the cable tensile flexibility term and the dynamic elastic flexibility term and taking the reciprocal. The calculation formula is as follows: in, Let be the dynamic equivalent stiffness of the i-th lifting point at time t, in N / m. It represents the overall resistance of the cable and hydraulic system at the lifting point to deformation. When the cable is in a slack state and the real-time tension is 0, the dynamic equivalent stiffness is taken as the preset minimum value 1e-6 N / m to avoid abnormal stiffness calculation results. When the calculated dynamic equivalent stiffness exceeds the preset reasonable range, the effective stiffness value of the previous control cycle is used for smooth transition. The reasonable range is set in advance according to the cable specifications and lifting conditions.

[0021] The second step is to reconstruct the dynamic force centroid using the real-time tension vector and absolute displacement vector of each suspension point, and orthogonally decouple the natural torsional deformation vector from the absolute displacement vector of each suspension point.

[0022] Multiply the real-time tension vector of each lifting point by its corresponding absolute displacement vector, sum the results, and divide by the sum of the real-time tension vectors of all lifting points to obtain the dynamic centroid of force. The calculation formula is as follows: in, Let t be the dynamic centroid of the hoisting system at time t, and let t be the three-dimensional coordinate vector in the construction independent coordinate system, with the unit being meters. Let t be the weighted average of the displacement of the hoisting points with the real-time tension of each hoisting point as the weight, which truly reflects the actual stress center of the connecting corridor under the load. When the sum of the real-time tension of all hoisting points is 0, the dynamic centroid is taken as the arithmetic mean of the absolute displacement vectors of all hoisting points.

[0023] Obtain the initial reference coordinates and initial centroid coordinates of the corresponding lifting points, as well as the three-dimensional rotation matrix output by the attitude sensor. Initial centroid coordinates The centroid of force, in meters, represents the initial static state of the connecting corridor. It is calculated using the dynamic centroid of force formula by substituting the initial tension at each lifting point and the initial reference coordinates into the formula. (Three-dimensional rotation matrix) Let t be the three-dimensional attitude rotation matrix of the connecting corridor body at time t. It is a 3×3 orthogonal identity matrix with no unit. It is collected and output in real time by a fiber optic gyroscope attitude sensor installed at the center of gravity of the connecting corridor body. The matrix elements correspond to the rotation angle transformation relationship of the connecting corridor around the X-axis, Y-axis and Z-axis of the construction independent coordinate system, which represents the rigid body rotation attitude of the connecting corridor body in space.

[0024] Multiplying the three-dimensional rotation matrix by the difference between the initial reference coordinates and the initial centroid coordinates of the lifting process, and adding the dynamic force centroid, yields the rigid body translational displacement vector of the corresponding lifting point. The calculation formula is as follows: The rigid body translation displacement vector is a three-dimensional coordinate vector in the construction independent coordinate system, with the unit being meters. It is the theoretical displacement coordinate of the i-th lifting point when the connecting corridor only undergoes rigid body translation and rotation. It only includes the displacement component brought about by the rigid body motion of the connecting corridor and does not include the deflection and torsional deformation of the connecting corridor itself.

[0025] Subtracting the rigid body translational displacement vector from the absolute displacement vector of the corresponding lifting point yields the natural deflection-torsional deformation vector of that point. The calculation formula is as follows: in, Let be the natural deflection and torsional deformation vector of the i-th lifting point at time t, which is a three-dimensional vector in the construction independent coordinate system, in meters. It is the difference between the actual absolute displacement of the lifting point and the theoretical displacement of the rigid body. It orthogonally decouples the displacement components brought about by the rigid body motion of the connecting corridor, and only retains the flexible deflection and torsional deformation of the connecting corridor itself caused by wind load and asymmetric load.

[0026] The third step is to generate a candidate command cluster containing multiple candidate velocity command vectors, and to construct a load surge penalty functional for evaluating each candidate velocity command vector based on the dynamic equivalent stiffness and natural torsional deformation vector of each suspension point.

[0027] Obtain the step size of the current control cycle and generate a candidate command cluster containing multiple candidate velocity command vectors. The candidate command cluster consists of multiple particles, each corresponding to a set containing all candidate velocity command vectors for each suspension point. The total number of particles is... This is a positive integer, ranging from 20 to 100. It can be adjusted based on the total number of lifting points and the computing power of the industrial controller; the more lifting points, the greater the corresponding number of particles. Candidate velocity command vector. The vector is composed of one-dimensional velocity values ​​along the cable lifting direction corresponding to the i-th lifting point, with units of meters per second. During the initial iteration, the candidate velocity command vectors of all particles are randomly initialized within the allowable lifting speed range of the connecting corridor hoisting. The allowable lifting speed range is 0 to 0.005 meters per second, which complies with the safety regulations for high-altitude hoisting operations.

[0028] Multiply the corresponding candidate velocity command vector by the step size, subtract the natural torsional deformation vector of the corresponding suspension point, and then calculate the modulus to obtain the counteracting displacement deviation. The calculation formula is as follows: The unit of counteracting displacement deviation is meters. It represents the difference between the theoretical displacement increment corresponding to the candidate speed command and the natural torsional deformation vector of the connecting corridor, reflecting the degree of counteracting of the natural torsional deformation of the connecting corridor by the candidate speed command.

[0029] Multiplying the dynamic equivalent stiffness of the corresponding suspension point by the counterbalancing displacement deviation yields the local surge load, calculated using the following formula: The unit of local surge load is Newtons (N), which is the local load increment at the i-th lifting point caused by the conflict between the candidate speed command and natural deformation. It conforms to the mechanical principle of Hooke's Law and can accurately reflect the risk of load surge caused by forced displacement.

[0030] The total parasitic internal force is obtained by summing the squares of the local surge loads at all lifting points. The calculation formula is as follows: The unit of total parasitic internal force is Newton squared, which represents the total parasitic strain energy generated by the forced displacement command in the entire hoisting system. The squared calculation can amplify the influence weight of the load surge at the hoisting point and prioritize avoiding the risk of local load exceeding the limit.

[0031] The normalized energy of the base is obtained by summing the squares of the magnitudes of the real-time tension vectors of all suspension points. The calculation formula is as follows: The unit of the base normalized energy is Newton squared, which represents the current foundation bearing capacity of the hoisting system. It is used to normalize the overall parasitic internal forces and eliminate numerical differences under different hoisting stages and different load levels. When the sum of the squares of the real-time tension moduli of all hoisting points is 0, the base normalized energy takes the preset minimum value 1e-9 to avoid the calculation error of dividing by 0.

[0032] Dividing the total parasitic internal force by the normalized energy of the base yields the load surge penalty functional for the corresponding candidate velocity command vector, calculated as follows: in, Let U be the load surge penalty functional corresponding to the set of candidate speed command vectors U. It is dimensionless. U is the set containing all candidate speed command vectors of the lifting points. The larger the penalty functional value, the greater the parasitic internal force of the system brought about by the candidate speed command, and the higher the risk of load surge.

[0033] The fourth step involves calculating the community information entropy of the candidate command community based on the load surge penalty functional, and then iteratively solving for the optimal speed command vector.

[0034] Obtain the first and second random numbers, and extract the optimal individual particle instruction and the global optimal instruction for the current generation of candidate instructions. First random number Second random number Uniformly distributed random numbers ranging from [0,1] are generated at each iteration to increase the randomness of the algorithm and avoid getting trapped in local optima. Individual particle optimal instructions. The candidate velocity command vector for a single particle when it obtains the minimum load surge penalty functional in the iteration history, and the swarm's globally optimal command. Let be the candidate velocity command vector corresponding to the minimum load surge penalty functional obtained in the iteration history of the entire candidate command swarm. In the initial iteration, the individual optimal command of a particle is initialized as the initial candidate command of that particle, and the swarm global optimal command is initialized as the initial candidate command with the minimum penalty functional among all particles.

[0035] The normalized particle superiority is obtained by taking the reciprocal of the load surge penalty functional of each particle in the candidate instruction swarm and dividing it by the sum of the reciprocals of the load surge penalty functionals of all particles. The calculation formula is as follows: in, denoted as the normalized particle superiority of the p-th particle in the k-th generation, it is dimensionless and ranges from [0,1]. Let $\mathbf{p}$ be the load surge penalty functional corresponding to the $p$-th particle in the $k$-th generation. The sum of the normalized particle superiority of all particles is $1$. When the load surge penalty functional of a particle is $0$, the normalized particle superiority of that particle is $1$, and the normalized particle superiority of the rest of the particles is $0$, to avoid the error of dividing by $0.

[0036] Multiplying the normalized particle dominance by the natural logarithm of the normalized particle dominance, summing the results over all particles, and taking the negative of the sum, yields the community information entropy. The formula is as follows: in, Let be the community information entropy of the k-th generation candidate instruction community, which is dimensionless and used to characterize the discreteness and diversity of particles in the community. ln is the natural logarithm function, which is the normalized particle superiority. When it is 0, The term is set to 0 to avoid the problem of meaningless logarithmic operations.

[0037] The decay inertia weight is obtained by exponentiation of the negative of the community information entropy with the natural constant as the base. The decay inertia weight is then multiplied by the current generation of candidate instructions to obtain the inertial evolution component. The calculation formula is as follows: in, This is the decay inertia weight, which is dimensionless and ranges from (0,1]. For the current generation candidate instruction of the p-th particle in the k-th generation, the inertial evolution component represents the particle's inheritance of its own historical motion state, and the global exploration capability and local development capability of the balance algorithm.

[0038] The difference between the normalized particle superiority score and the first random number, multiplied by the difference between the individual particle's optimal instruction and the current generation's candidate instructions, yields the individual cognitive optimization component. The calculation formula is as follows: The individual cognitive optimization component is the velocity component of a particle learning from its own historical best solution, and 1-W_p is the individual learning weight. The lower the particle's superiority, the larger the individual learning weight, and the more the particle tends to learn from its own historical best solution.

[0039] Multiplying the normalized particle superiority by the second random number and the difference between the swarm's global optimal instruction and the current generation's candidate instructions yields the social learning optimization component, calculated as follows: The social learning optimization component is the velocity component of particles learning from the global optimal solution of the population. As the social learning weight, the higher the particle's superiority, the greater the social learning weight, and the more the particle tends to learn from the global optimal solution of the group.

[0040] Adding the inertial evolution component, the individual cognitive optimization component, and the social learning optimization component yields the updated next-generation candidate instructions. The calculation formula is as follows: in, For the updated candidate instructions of the p-th particle in the (k+1)-th generation, after each iteration, the load surge penalty functional of all particles is recalculated. If the penalty functional of the current generation particle is less than the historical individual optimal penalty functional of the particle, the individual optimal instruction of the particle is updated; if the minimum penalty functional of the current generation population is less than the historical global optimal penalty functional, the global optimal instruction of the population is updated.

[0041] During the iteration process, the convergence criteria are continuously checked. These criteria include reaching a preset maximum number of iterations, the global optimal penalty functional showing no decrease within a preset number of consecutive iterations, and the community information entropy being less than a preset convergence threshold. The iteration terminates when any one of these criteria is met. The preset maximum number of iterations ranges from 50 to 200, the preset number of iterations showing no decrease ranges from 10 to 20, and the preset convergence threshold ranges from 0.01 to 0.1, which can be adjusted according to the control accuracy requirements of the hoisting operation. After convergence, the globally optimal command that minimizes the load surge penalty functional is output, serving as the optimal speed command vector for each hoisting point. .

[0042] The fifth step is to obtain real-time wind speed to calculate the total venturi wind driving force, and then combine the natural torsional deformation vector of each suspension point with the optimal speed command vector to generate the target command tension.

[0043] Obtain air density and windward projected area, as well as real-time wind speed and wind direction angle. Air density The unit is kg / m³, and the value under standard conditions is 1.225 kg / m³. It can be corrected using the ideal gas law based on the altitude and ambient temperature. Real-time wind speed. The real-time wind speed at the hoisting location of the connecting corridor at time t is measured in meters per second. This data is collected in real-time by an anemometer installed on the windward side of the corridor. The collected data undergoes low-pass filtering with a cutoff frequency of 1Hz to suppress high-frequency disturbances caused by gusts. The wind direction angle is the angle between the real-time wind direction and the X-axis of the construction-independent coordinate system. The corresponding pre-stored windward projected area is used based on the wind direction angle. At the same time, it is multiplied by the wind speed amplification factor of the canyon wind field, which ranges from 1.1 to 1.5 and is determined by the canyon space dimensions at the hoisting site.

[0044] Multiplying half the area, air density, windward projected area, and the square of the real-time wind speed together, we obtain the Venturi total wind driving force. The calculation formula is as follows: in, Let t be the total Venturi wind driving force on the connecting corridor at time t, in Newtons. This is consistent with the calculation principle of wind load in fluid mechanics and characterizes the total aerodynamic load on the connecting corridor in the canyon wind field.

[0045] Multiplying the dynamic equivalent stiffness of the corresponding lifting point, the optimal velocity command vector, and the step size together yields the rigid displacement driving force, calculated using the following formula: The unit of rigid displacement driving force is Newton. It is the driving force required to achieve the displacement increment corresponding to the optimal speed command. It conforms to Hooke's Law and can accurately characterize the tension increment required for the lifting point to move according to the optimal speed command.

[0046] Divide the magnitude of the natural torsional deformation vector at the corresponding suspension point by the sum of the magnitudes of the natural torsional deformation vectors at all suspension points to obtain the deformation weight ratio. The calculation formula is as follows: The deformation weight ratio is a unitless parameter with a value range of [0,1]. It represents the degree of deflection and torsion deformation of the connecting corridor at the i-th suspension point. When the sum of the natural deflection and torsion deformation vector magnitudes of all suspension points is 0, the deformation weight ratio of all suspension points is taken as the average value 1 / n to avoid the error of dividing by 0.

[0047] Multiply the total Venturi wind-driven force by the deformation weight ratio, and then multiply by the ratio of the real-time tension vector at the corresponding suspension point to its own modulus to obtain the aerodynamic compensation force. The calculation formula is as follows: The unit of aerodynamic compensation force is Newtons (N), which represents the wind load compensation force that the i-th lifting point needs to bear. The unit direction vector of the real-time tension vector at the i-th suspension point is used to decompose the total wind-driven force along the cable axis to ensure that the direction of the compensation force is consistent with the direction of the cable tension. When the magnitude of the real-time tension vector is 0, the unit direction vector is taken as the vertically upward unit vector.

[0048] The target command tension at the corresponding lifting point is obtained by adding the real-time tension vector, the rigid displacement driving force, and the aerodynamic compensation force at the corresponding lifting point. The calculation formula is as follows: in, Let t be the target command tension at the i-th lifting point at time t, in Newtons. It is the sum of the current real-time tension, rigid displacement driving force, and aerodynamic compensation force. It includes both the tension increment required to achieve the optimal lifting speed and the compensation component for wind load disturbance. The calculated target command tension must be limited to within 50% of the cable's rated breaking tension to comply with the safety regulations for lifting operations.

[0049] The sixth step is to substitute the target command pull force into the fluid control equation and solve it inversely to obtain the mechanical opening area used to control the corresponding proportional valve.

[0050] Obtain the hydraulic oil volume and pump station pressure for the corresponding hydraulic cylinder. Hydraulic oil volume Let be the volume of hydraulic oil in the rodless chamber of the hydraulic cylinder at time t, in cubic meters. This volume is calculated by multiplying the real-time stroke of the hydraulic cylinder by its effective area, and it changes in real-time with the cylinder's stroke. Pump station pressure. The output oil supply pressure of the hydraulic pump station at time t is measured in Pascals and is collected in real time by a pressure sensor at the pump station outlet.

[0051] Multiplying the effective area of ​​the hydraulic cylinder by the modulus of the optimal speed command vector at the corresponding lifting point yields the basic volumetric flow rate. The calculation formula is as follows: The unit of basic volumetric flow rate is cubic meters per second. The basic flow rate of hydraulic oil required to achieve the optimal speed command conforms to the correspondence between the hydraulic cylinder movement speed and the input flow rate, that is, the flow rate is equal to the effective area multiplied by the piston movement speed.

[0052] The mold length is obtained by subtracting the target command tension from the real-time tension vector at the corresponding lifting point. This mold length is then divided by the product of the step length and the effective area of ​​the hydraulic cylinder to obtain the required force change rate. The calculation formula is as follows: The required force change rate is measured in Pascals per second (Pa), representing the pressure change rate of the rodless chamber of the hydraulic cylinder per unit time. The numerator is the modulus of the difference between the target command tension and the current real-time tension, i.e., the amount of tension change to be achieved within the current control cycle. The denominator is the product of the control cycle step size and the effective area of ​​the hydraulic cylinder, converting the amount of tension change into the pressure change rate.

[0053] Divide the hydraulic fluid volume by the fluid's elastic modulus, then multiply by the required force change rate to obtain the compressible compensation flow rate. The calculation formula is as follows: The unit of compressible compensation flow rate is cubic meters per second. It is the additional flow rate required to compensate for the compressibility of hydraulic oil. The volumetric compliance of hydraulic oil characterizes its ability to deform in volume under pressure changes. Multiplying the volumetric compliance by the pressure change rate yields the volume change rate caused by the compressibility of the hydraulic oil, which is the additional flow rate required.

[0054] Divide the modulus of the target tensile force by the effective area of ​​the hydraulic cylinder to obtain the load pressure. Calculate the absolute value of the difference between the pump station pressure and the load pressure. Multiply the absolute value of the difference by two, divide by the oil density, and take the arithmetic square root to obtain the pressure differential velocity term. The calculation formula is as follows: The pressure differential velocity term is measured in meters per second and represents the velocity characteristic term corresponding to the pressure difference across the proportional valve orifice. The load pressure of the rodless chamber of the hydraulic cylinder is expressed in Pascals. The absolute value of the difference between the pump station pressure and the load pressure is the pressure difference before and after the valve port. This is consistent with the calculation logic of the pressure difference term in the thin-walled orifice flow formula. When the absolute value of the pressure difference is 0, the pressure difference flow rate term takes the preset minimum value 1e-9 to avoid the calculation error of dividing by 0.

[0055] Add the basic volumetric flow rate to the compressible compensated flow rate, and divide by the product of the flow coefficient and the differential pressure velocity term to obtain the mechanical opening area used to control the corresponding proportional valve. The calculation formula is as follows: in, Let be the mechanical opening area of ​​the proportional valve corresponding to the i-th lifting point at time t, in square meters. The numerator is the sum of the basic volumetric flow rate and the compressible compensation flow rate, which is the total flow rate that needs to pass through the valve orifice during the control cycle. The denominator is the product of the flow coefficient and the differential pressure velocity term, which conforms to the thin-walled orifice flow rate formula. The required mechanical opening area of ​​the valve orifice can be obtained by inverse solving the total flow rate.

[0056] Based on the correspondence between the rated opening area of ​​the proportional valve and the rated control electrical signal, the mechanical opening area is converted into the control electrical signal of the proportional valve. The control electrical signal is a standard current signal of 4 to 20 mA or a standard voltage signal of 0 to 10 V. The industrial controller outputs the control electrical signal to the servo amplifier of the proportional valve, driving the proportional valve to adjust the valve opening, thereby realizing closed-loop control of the hydraulic cylinder.

[0057] The seventh step is to calculate the background natural vibration period based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, and then combine the community information entropy to obtain the control step size of the next period.

[0058] The preset values ​​for gravitational acceleration and pi are: g = 9.8 m / s² and pi = 3.1415926535.

[0059] The equivalent mass of the system is obtained by summing the magnitudes of the real-time tension vectors at all suspension points and dividing by the gravitational acceleration. The calculation formula is as follows: The unit of equivalent mass of the system is kilogram, which is the overall equivalent mass of the connecting corridor hoisting system and conforms to the conversion relationship between gravity and mass.

[0060] The overall system stiffness is obtained by summing the dynamic equivalent stiffness of each suspension point. The calculation formula is as follows: The unit of overall system stiffness is Newton-meter (N / m), which is the sum of the dynamic equivalent stiffness of all lifting points and characterizes the overall ability of the entire lifting system to resist deformation.

[0061] Divide the system's equivalent mass by the overall system stiffness and take the arithmetic square root, then multiply by pi to obtain the base natural period. The calculation formula is as follows: The unit of the natural period is seconds. It is the natural period of the hoisting system and conforms to the calculation logic of the natural period of a single degree of freedom system. It can reflect the inherent vibration characteristics of the hoisting system. When the overall stiffness of the system is 0, the natural period is set to the preset maximum value of 0.1 seconds to avoid the calculation error of dividing by 0.

[0062] Dividing 1 by the community information entropy and taking its negative value, then using this as the exponent of the natural constant to obtain the dynamic scaling term, and subtracting the dynamic scaling term from 1, yields the adaptive adjustment coefficient. The calculation formula is as follows: The adaptive adjustment coefficient is a unitless parameter, with a value range of [0,1]. The larger the community information entropy, the higher the dispersion of the particle swarm, indicating that the optimization process has not converged. The smaller the adaptive adjustment coefficient, the smaller the control step size in the next cycle, thus improving control accuracy. The smaller the community information entropy, the more the particle swarm tends to converge, indicating that the optimization process is stable. The larger the adaptive adjustment coefficient, the larger the control step size in the next cycle, thus improving control response speed. When the community information entropy is 0, the adaptive adjustment coefficient is 1 to avoid the error of dividing by 0.

[0063] Multiplying the natural period by the adaptive adjustment coefficient yields the control step size for the next period. The calculation formula is as follows: in, The step size for the next control cycle is in seconds. It is the product of the natural oscillation period and the adaptive adjustment coefficient. The calculated control step size must be limited to the preset upper and lower limits, with an upper limit of 0.1 seconds and a lower limit of 0.01 seconds. If the step size exceeds the range, the corresponding boundary value is taken to avoid the control becoming unstable due to an excessively large step size or the industrial controller becoming overloaded due to an excessively small step size.

[0064] After all operations in the current control cycle are completed, the industrial controller latches the control step size for the next cycle into the register, enters the next control cycle, and repeats all the above steps until the hoisting operation of the connecting corridor is completed.

[0065] Example 2: A posture detection, synchronous lifting, and fine-tuning control system for connecting corridor hoisting, realizing the posture detection, synchronous lifting, and fine-tuning control process as described in any one of the embodiments, including: The stiffness calculation module is used to obtain the cable free length, absolute displacement vector and real-time tension vector of each suspension point within the current control cycle, and adaptively calculate the dynamic equivalent stiffness of each suspension point. The deformation decoupling module is used to reconstruct the dynamic force centroid using the real-time tension vector and the absolute displacement vector of each of the suspension points, and to orthogonally decouple the natural torsional deformation vector from the absolute displacement vector of each of the suspension points. A functional construction module is used to generate a candidate command cluster containing multiple candidate velocity command vectors, and to construct a load surge penalty functional for evaluating each candidate velocity command vector based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points. The optimization module is used to calculate the community information entropy of the candidate command community based on the load surge penalty functional, and to calculate the optimal speed command vector through optimization iteration. The tension generation module is used to obtain real-time wind speed to calculate the total Venturi wind driving force, and to generate the target command tension by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. The valve port control module is used to substitute the target command pull force into the fluid control equation and solve it inversely to obtain the mechanical opening area used to control the corresponding proportional valve. The step size evolution module is used to calculate the background natural vibration period based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, and to obtain the control step size of the next period by combining the community information entropy.

[0066] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. The attitude detection, synchronous lifting, and fine-tuning control technology for the hoisting of a connecting corridor is applied to a control system that includes multi-point suspended cables and hydraulic actuators with proportional valves, and is executed within the current control cycle. Its characteristics are as follows: include: The free length of the cable, the absolute displacement vector and the real-time tension vector of each suspension point are obtained, and the dynamic equivalent stiffness of each suspension point is adaptively calculated. The dynamic force centroid is reconstructed using the real-time tension vector and the absolute displacement vector of each of the aforementioned suspension points, and the natural deflection vector is orthogonally decoupled from the absolute displacement vector of each of the aforementioned suspension points. A candidate command cluster containing multiple candidate velocity command vectors is generated. Based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points, a load surge penalty functional for evaluating each of the candidate velocity command vectors is constructed. The community information entropy of the candidate command community is calculated based on the load surge penalty functional, and the optimal speed command vector is calculated through optimization iteration. Real-time wind speed is obtained to calculate the total Venturi wind driving force, and the target command tension is generated by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. Substituting the target command force into the fluid control equation, the mechanical opening area used to control the corresponding proportional valve is obtained by inverse solution; Based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, the natural vibration period is calculated, and the control step size of the next period is obtained by combining the community information entropy.

2. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 1, characterized in that, Obtain the cable free length, absolute displacement vector, and real-time tension vector at each suspension point, and adaptively calculate the dynamic equivalent stiffness of each suspension point, including: Obtain the cable elastic modulus and cable cross-sectional area of ​​the multi-point suspended cable, as well as the absolute displacement vector and real-time tension vector of the previous control cycle. The cable tensile flexibility term is obtained by dividing the free length of the cable at each suspension point by the product of the cable elastic modulus and the cable cross-sectional area. Calculate the absolute value of the difference between the absolute displacement vector of each of the lifting points and the absolute displacement vector of the previous control cycle to obtain the displacement difference modulus; Calculate the absolute value of the difference between the real-time tension vector of each of the aforementioned lifting points and the real-time tension vector of the previous control cycle to obtain the tensile force differential modulus. Dividing the displacement differential modulus by the tensile force differential modulus yields the dynamic elastic compliance term; The dynamic equivalent stiffness of each suspension point is obtained by adding the cable tensile flexibility term and the dynamic elastic flexibility term and taking the reciprocal.

3. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 2, characterized in that, The dynamic force centroid is reconstructed using the real-time tension vector and absolute displacement vector of each of the aforementioned suspension points, and the natural deflection-torsional deformation vector is orthogonally decoupled from the absolute displacement vector of each of the aforementioned suspension points, including: The dynamic force centroid is obtained by multiplying the real-time tension vector of each suspension point with the corresponding absolute displacement vector, summing the results, and dividing by the sum of the real-time tension vectors of all suspension points. Obtain the initial reference coordinates and initial centroid coordinates of the corresponding lifting point, as well as the three-dimensional rotation matrix output by the attitude sensor; Multiply the three-dimensional rotation matrix by the difference between the initial lifting reference coordinates and the initial lifting centroid coordinates, and add the dynamic force centroid to obtain the rigid body translational displacement vector of the corresponding lifting point. Subtracting the rigid body translational displacement vector from the absolute displacement vector of the corresponding suspension point yields the natural torsional deformation vector of the corresponding suspension point.

4. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 3, characterized in that, Based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the aforementioned suspension points, a load surge penalty functional for evaluating each of the aforementioned candidate velocity command vectors is constructed, including: Obtain the step size of the current control cycle, multiply the corresponding candidate speed command vector by the step size, subtract the natural torsional deformation vector of the corresponding suspension point, and then calculate the modulus to obtain the counteracting displacement deviation. Multiply the dynamic equivalent stiffness of the corresponding suspension point by the counteracting displacement deviation to obtain the local surge load; The total parasitic internal force is obtained by summing the squares of the local surge loads at all suspension points. The normalized energy of the base is obtained by summing the squares of the magnitudes of the real-time tension vectors of all the suspension points. Dividing the total parasitic internal force by the normalized energy of the base yields the load surge penalty functional of the corresponding candidate velocity command vector.

5. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 4, characterized in that, The community information entropy of the candidate command cluster is calculated based on the load surge penalty functional, and the optimal velocity command vector is obtained through optimization iteration, including: The candidate instruction cluster consists of multiple particles. A first random number and a second random number are obtained, and the optimal individual instruction and the global optimal instruction of the current generation of candidate instructions are extracted. The normalized particle superiority is obtained by taking the reciprocal of the load surge penalty functional of each particle in the candidate instruction cluster and dividing it by the sum of the reciprocals of the load surge penalty functional of all particles. Multiply the normalized particle superiority by the natural logarithm of the normalized particle superiority, sum over all particles, and take the opposite number to obtain the community information entropy. The decay inertia weight is obtained by exponentiation of the community information entropy with the natural constant as the base and the negative number as the exponent. The decay inertia weight is then multiplied by the current generation of candidate instructions to obtain the inertial evolution component. The difference between the normalized particle superiority and the first random number, and the difference between the individual particle optimal instruction and the current generation candidate instruction, are multiplied by the difference to obtain the individual cognitive optimization component. Multiply the normalized particle superiority by the second random number and the difference between the population global optimal instruction and the current generation candidate instruction to obtain the social learning optimization component; The inertial evolution component, the individual cognitive optimization component, and the social learning optimization component are added together to obtain the updated next-generation candidate instruction. After iterative convergence, the globally optimal speed instruction that minimizes the load surge penalty functional is output as the optimal speed instruction vector for each of the suspension points.

6. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 5, characterized in that, Real-time wind speed is acquired to calculate the total Venturi wind driving force. The target command tension is generated by combining the natural torsional deformation vectors of each suspension point with the optimal velocity command vector, including: Obtain the air density and windward projected area, and multiply half of the air density, the windward projected area, and the square of the real-time wind speed to obtain the Venturi total wind driving force. The dynamic equivalent stiffness of the corresponding suspension point, the optimal speed command vector, and the step size are multiplied together to obtain the rigid displacement driving force. Divide the magnitude of the natural torsional deformation vector of the corresponding suspension point by the sum of the magnitudes of the natural torsional deformation vectors of all suspension points to obtain the deformation weight ratio; Multiply the total Venturi wind driving force by the deformation weight ratio, and then multiply by the ratio of the real-time tension vector of the corresponding suspension point to its own modulus to obtain the aerodynamic compensation force. The target command tension of the corresponding lifting point is obtained by adding the real-time tension vector, the rigid displacement driving force, and the pneumatic compensation sharing force at the corresponding lifting point.

7. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 6, characterized in that, Substituting the target command force into the fluid control equation, the inverse solution yields the mechanical opening area used to control the corresponding proportional valve, including: The hydraulic actuator is a hydraulic cylinder. The effective area of ​​the hydraulic cylinder, the volume of the hydraulic oil, the elastic modulus of the oil, the pump station pressure, the oil density and the flow coefficient are obtained. Multiply the effective area of ​​the hydraulic cylinder by the modulus of the optimal speed command vector at the corresponding lifting point to obtain the basic volumetric flow rate; The modulus is obtained by subtracting the target command tension from the real-time tension vector at the corresponding lifting point, and then dividing the modulus by the product of the step length and the effective area of ​​the hydraulic cylinder to obtain the required force change rate. Divide the hydraulic fluid volume by the fluid elastic modulus, and then multiply by the required force change rate to obtain the compressible compensation flow rate. The load pressure is obtained by dividing the modulus of the target command tension by the effective area of ​​the hydraulic cylinder. The absolute value of the difference between the pump station pressure and the load pressure is calculated. The absolute value of the difference is multiplied by two, divided by the oil density, and the arithmetic square root is obtained to obtain the pressure difference velocity term. The basic volumetric flow rate is added to the compressible compensated flow rate and divided by the product of the flow coefficient and the differential pressure velocity term to obtain the mechanical opening area used to control the corresponding proportional valve.

8. The posture detection, synchronous lifting, and fine-tuning control process for the hoisting of the connecting corridor according to claim 7, characterized in that, Based on the real-time tension vectors of all suspension points and the dynamic equivalent stiffness of each suspension point, the natural vibration period is calculated, and the control step size for the next period is obtained by combining the community information entropy, including: Preset gravitational acceleration and pi; The equivalent mass of the system is obtained by summing the magnitudes of the real-time tension vectors at all suspension points and dividing by the gravitational acceleration. The overall system stiffness is obtained by summing the dynamic equivalent stiffness of each of the aforementioned suspension points; The system's equivalent mass is divided by the system's overall stiffness, and the arithmetic square root is obtained. This square root is then multiplied by the value of pi to obtain the system's natural period. Divide one by the community information entropy and take the opposite number, then use it as the exponent of the natural constant to obtain the dynamic scaling term. Subtract the dynamic scaling term from one to obtain the adaptive adjustment coefficient. The control step size for the next cycle is obtained by multiplying the natural vibration period by the adaptive adjustment coefficient.

9. A posture detection, synchronous lifting, and fine-tuning control system for the hoisting of a connecting corridor, wherein the posture detection, synchronous lifting, and fine-tuning control process for the hoisting of a connecting corridor as described in any one of claims 1-8 is characterized in that, include: The stiffness calculation module is used to obtain the cable free length, absolute displacement vector and real-time tension vector of each suspension point within the current control cycle, and adaptively calculate the dynamic equivalent stiffness of each suspension point. The deformation decoupling module is used to reconstruct the dynamic force centroid using the real-time tension vector and the absolute displacement vector of each of the suspension points, and to orthogonally decouple the natural torsional deformation vector from the absolute displacement vector of each of the suspension points. A functional construction module is used to generate a candidate command cluster containing multiple candidate velocity command vectors, and to construct a load surge penalty functional for evaluating each candidate velocity command vector based on the dynamic equivalent stiffness and natural torsional deformation vector of each of the suspension points. The optimization module is used to calculate the community information entropy of the candidate command community based on the load surge penalty functional, and to calculate the optimal speed command vector through optimization iteration. The tension generation module is used to obtain real-time wind speed to calculate the total Venturi wind driving force, and to generate the target command tension by combining the natural torsional deformation vector of each suspension point with the optimal speed command vector. The valve port control module is used to substitute the target command pull force into the fluid control equation and solve it inversely to obtain the mechanical opening area used to control the corresponding proportional valve. The step size evolution module is used to calculate the background natural vibration period based on the real-time tension vector of all suspension points and the dynamic equivalent stiffness of each suspension point, and to obtain the control step size of the next period by combining the community information entropy.

Images