An industrial equipment installation precision control method and system based on BIM and laser fusion positioning

Abstract

The application relates to the field of industrial process computer monitoring and control, and discloses an industrial equipment installation precision control method and system based on BIM and laser fusion positioning, which comprises the following steps: extracting a process topological logic node in a topological constraint reference model, establishing a pose true value coordinate and an environmental thermal expansion coefficient tensor; obtaining a multi-source space feedback signal stream of a controlled object and extracting a geometric characteristic parameter, and mapping the geometric characteristic parameter and the process topological logic node in space-time; inhibiting a discrete pose vector by using a topological consistency arbitration logic to construct a normalized topological deviation field, and generating a trajectory correction instruction according to the topological deviation field and sending the trajectory correction instruction to an adjusting unit to drive the pose regulation and control of the controlled object. The application solves the closed-loop instability problem in the precise installation process, dynamically synchronizes the control reference and the physical entity state, and enhances the stability of the system in a harsh working environment.

Description

Technical Field

[0001] This invention belongs to the field of industrial process computer supervision and control technology, and in particular relates to a method and system for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning. Background Technology

[0002] The installation of industrial equipment is currently at a critical stage of transformation from traditional manual centering to digital and automated closed-loop supervision and control. In the installation of large-scale precision industrial equipment, computer supervision systems establish geometric logic truth values ​​by calling building information models and use laser scanning systems to collect physical coordinate flows of characteristic parts of the equipment. The spatial deviation between the two is compared to drive the actuator to complete the positioning. This supervision mode has high reliability when dealing with small-sized, high-rigidity components and has become the mainstream technical path in the field of precision assembly. However, with the popularization of ultra-large and precision equipment such as semiconductor lithography machine bases, core components of nuclear islands, and flexible assembly lines for aero-engines, the physical environment of the installation site is becoming more complex. Affected by uncontrolled variables such as gravitational deformation, environmental stress disturbances, and micro-vibrations of the foundation, the physical entity of the equipment to be installed exhibits non-rigid micro-deformation characteristics. Existing computer supervision strategies are based on rigid body dynamics assumptions and only implement point-to-point geometric coincidence matching. Due to the lack of a decoupling mechanism for global position deviation and local structural distortion in the pose sampling data, the supervision system is unable to accurately extract pure deviation quantities in non-ideal environments.

[0003] Digital control logic often exhibits insufficient robustness in complex operating conditions. For example, Chinese invention patent application CN121504807A discloses a method for detailed control of the installation of electromechanical equipment in hydropower engineering based on 3D laser scanning technology. This method uses the RANSAC algorithm to fit a target sphere model and compares the coordinate deviation with that of a virtual target sphere. While this technique improves sampling accuracy through feature fitting, it still relies on the rigid presupposition of a fixed geometric shape, treating the equipment as an indeformable ideal geometric body. In ultra-large precision installation scenarios, the equipment may experience sub-millimeter-level local distortions due to gravity or thermal stress. The logic based on global feature fitting leads to a fundamental mismatch between the sampled point cloud and the model's topological nodes. The monitoring system cannot identify the impact of local distortions on global pose calculation, inducing overshoot of the adjustment pulse. Or path oscillations, existing technologies lack physical modeling of environmental thermal instability, and the control reference drifts under varying temperature conditions; when large equipment experiences local micro-bending at the level of 0.05mm, the coordinates of physical points obtained by laser sampling mismatch with the topological nodes in the building information model, leading to the introduction of discrete noise containing structural stress characteristics into the control feedback loop. Simply increasing the sampling frequency or calculation step size cannot suppress such logical inaccuracies, but instead causes overshoot of the correction pulse, resulting in impacts to precision interfaces or structural damage. The industry has tried to reduce errors by adding physical auxiliary supports or preset static deviation fields, but these methods not only increase process costs, but also cannot cope with the dynamically evolving environmental thermal instability problems during installation, creating an irreconcilable technical constraint between installation accuracy and system robustness.

[0004] Therefore, the technical problem to be solved by this invention is how to establish an industrial process control logic that can achieve global deviation decoupling and has the ability to adaptively correct the reference, so as to ensure the closed-loop installation accuracy of large non-rigid equipment under complex working conditions. Summary of the Invention

[0005] This invention aims to solve the problems of inaccurate monitoring logic and non-convergence of control paths caused by non-rigid deformation during the installation of large precision equipment.

[0006] In this technical solution, an industrial equipment installation accuracy control method based on BIM and laser fusion positioning includes the following steps:

[0007] Step 101: Extract the process topology logic nodes that represent the key assembly positions of the controlled object in the topology constraint reference model, and establish the true pose coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor used to compensate for thermal instability.

[0008] Step 102: Real-time acquisition of multi-source spatial feedback signal streams for the physical entity of the controlled object, and extraction of geometric feature parameters covering edge contours, feature apertures and surface curvature information from the multi-source spatial feedback signal streams;

[0009] Step 103: Perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and identify and suppress discrete pose vectors in geometric feature parameters that exceed the preset deviation threshold due to dust occlusion or multiple reflections through topology consistency arbitration logic integrated into the computer supervision system, thereby constructing a normalized topology deviation field.

[0010] Step 104: Based on the normalized topological deviation field, rotation and translation operators are extracted through nonlinear calculation to generate trajectory correction instructions. The correction adjustment signal containing installation displacement compensation information is sent to the end adjustment unit. The controlled object is driven to complete precise pose control through digital closed-loop logic supervision.

[0011] Preferably, the topology consistency arbitration logic in step 103 includes: calculating the deviation vector of each geometric feature parameter relative to the process topology logic node; when the cross-correlation coefficient between the deviation vector and the deviation trend of neighboring nodes is lower than a preset logic threshold, automatically reducing the weight parameter of the node, and using the logic integrity of the remaining nodes to maintain the output stability of the topology deviation field.

[0012] Preferably, the trajectory correction instruction generated in step 104 further includes: generating a six-degree-of-freedom attitude adjustment vector based on rotation and translation operators, calculating the deformation residual between geometric feature parameters and pose true coordinates; when the deformation residual exceeds a preset safety threshold, outputting a warning signal for the controlled object and triggering a logical suspension action in the installation process.

[0013] Preferably, the topology consistency arbitration logic further includes: determining the proportion of effective data points in the multi-source spatial feedback signal stream to determine the signal sampling quality; when the signal sampling quality is less than 60%, switching the supervision mode from the global image mode to the key target alignment mode, and generating a correction adjustment signal using the local topology information of the preset physical target to achieve precision control.

[0014] Preferably, the multi-source spatial feedback signal stream in step 102 is real-time point cloud data generated by non-contact detection of the controlled object using a high-frequency laser scanning device; the extraction process of geometric feature parameters includes: denoising the real-time point cloud data, and locking the geometric centroid and axis vector information of the controlled object based on the curvature change point identification algorithm.

[0015] Preferably, the correction adjustment signal is used to drive the end effector unit to generate physical compensation displacement in three dimensions, so as to reduce the real-time pose deviation of the controlled object relative to the topological constraint reference model; the end effector unit includes a multi-axis drive system composed of a hydraulic drive mechanism or a servo module, and the multi-axis drive system completes pose fine adjustment in response to the correction adjustment signal.

[0016] Preferably, when extracting process topology logic nodes in step 101, a topology connection matrix between each process topology logic node is established simultaneously; the topology connection matrix defines the physical assembly constraint relationship of different interface components inside the controlled object; the topology consistency arbitration logic performs logical verification on the pose evolution process of geometric feature parameters based on the topology connection matrix.

[0017] Preferably, in step 103, when establishing the normalized topological deviation field, a nonlinear least squares fitting algorithm is used to iteratively optimize the geometric feature parameters to minimize the sum of squared Euclidean distances between the geometric feature parameters and the true pose coordinates; the normalization process includes mapping the dynamic pose deviation of the controlled object to a preset zero-bias reference space to eliminate system background noise interference.

[0018] Preferably, the precision control method is accomplished through a monitoring program deployed in an industrial control terminal; the monitoring program outputs dynamic correction pulses by comparing the transient installation trajectory of the controlled object with the theoretical path preset by the topological constraint reference model in real time; the industrial control terminal has a real-time communication interface for transmitting trajectory correction instructions to the automated drive module of the controlled object.

[0019] An industrial equipment installation accuracy control system based on BIM and laser fusion positioning includes:

[0020] The topology reference extraction module is used to extract process topology logic nodes that represent key assembly positions of the controlled object from the topology constraint reference model, and to establish the pose true coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor.

[0021] The feedback signal acquisition module is used to acquire multi-source spatial feedback signal streams for the controlled object in real time and extract geometric feature parameters from the multi-source spatial feedback signal streams.

[0022] The topology arbitration calculation module is used to perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and to identify and suppress discrete pose vectors in geometric feature parameters through topology consistency arbitration logic to generate a normalized topology deviation field.

[0023] The correction and adjustment output module is used to extract rotation and translation operators based on the normalized topological deviation field to generate trajectory correction instructions, and send the correction and adjustment signal containing installation displacement compensation information to the end adjustment unit to drive the controlled object to complete the pose control.

[0024] Compared with existing technologies, the industrial equipment installation accuracy control method based on BIM and laser fusion positioning of this invention has the following advantages:

[0025] 1. In the precision control of industrial equipment installation, the geometric topology of the building information model is transformed into a logical node mesh with physical stiffness constraints. Combined with the physical point cloud of the equipment's characteristic parts, a weighted deviation energy function is calculated. This achieves logical decoupling between the global pose transformation operator and the local deformation residual term. This mechanism changes the traditional installation method that relies solely on point-to-point geometric matching. The computer monitoring system, in the process of minimizing the energy function, automatically removes non-rigid micro-deformation interference caused by gravity, environmental stress, or uneven support through a weight matrix. This method of purifying and extracting pose deviations at the control logic level effectively avoids erroneous compensation actions by the installation actuator for local micro-deformations, eliminates control command overshoot and repeated path oscillations, and ensures that the positioning accuracy of large industrial equipment under complex stress conditions remains stable within 0.01mm.

[0026] 2. A topology consistency arbitration mechanism is introduced. By monitoring the deviation trend between physical sampling points and adjacent logical nodes, the node weights are dynamically adjusted. The overall logical integrity of the topology structure is used to compensate for the lack or distortion of local data. In industrial environments where dust obstruction and interference from metal reflective surfaces cause point cloud quality to drop to 60%, the arbitrator automatically identifies and suppresses discrete pose vectors containing high-frequency noise, ensuring that the closed-loop adjustment pulses output by the monitoring system maintain spatiotemporal consistency. This control strategy based on logical redundancy verification enhances the system's robustness to harsh operating environments, ensures the smooth convergence of automated installation processes, and improves the convergence speed of the overall control process by more than 40%.

[0027] 3. By integrating the thermal expansion coefficient tensor into the building information model and combining it with real-time feedback from environmental sensors to correct the theoretical coordinates of logical nodes online, an environmental thermal instability compensation logic is constructed. This method enables the computer monitoring system to sense and counteract the reference drift caused by the thermal expansion and contraction of ultra-large equipment, and achieve dynamic synchronization between the control reference and the physical entity state. This collaborative mechanism, which directly transforms environmental physical variables into control logic reference correction, solves the long-standing topological inaccuracy contradiction between the static true value of the model and the transient measured value of the physics, and enables the system to still have adaptive submicron-level monitoring and control capabilities under variable temperature conditions. Attached Figure Description

[0028] Figure 1 This is a flowchart of the precision installation control process for the fusion of BIM topology constraints and laser pose in this invention.

[0029] Figure 2 This is a diagram of the closed-loop monitoring system architecture of the present invention, which features topology consistency arbitration and environmental adaptive compensation. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0031] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0032] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0033] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0034] A method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning includes the following steps:

[0035] Step 101: Extract the process topology logic nodes that represent the key assembly positions of the controlled object in the topology constraint reference model, and establish the true pose coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor used to compensate for thermal instability.

[0036] Step 102: Real-time acquisition of multi-source spatial feedback signal streams for the physical entity of the controlled object, and extraction of geometric feature parameters covering edge contours, feature apertures and surface curvature information from the multi-source spatial feedback signal streams;

[0037] Step 103: Perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and identify and suppress discrete pose vectors in geometric feature parameters that exceed the preset deviation threshold due to dust occlusion or multiple reflections through topology consistency arbitration logic integrated into the computer supervision system, thereby constructing a normalized topology deviation field.

[0038] Step 104: Based on the normalized topological deviation field, rotation and translation operators are extracted through nonlinear calculation to generate trajectory correction instructions. The correction adjustment signal containing installation displacement compensation information is sent to the end adjustment unit. The controlled object is driven to complete precise pose control through digital closed-loop logic supervision.

[0039] Preferably, the topology consistency arbitration logic in step 103 includes: calculating the deviation vector of each geometric feature parameter relative to the process topology logic node; when the cross-correlation coefficient between the deviation vector and the deviation trend of neighboring nodes is lower than a preset logic threshold, automatically reducing the weight parameter of the node, and using the logic integrity of the remaining nodes to maintain the output stability of the topology deviation field.

[0040] Preferably, the trajectory correction instruction generated in step 104 further includes: generating a six-degree-of-freedom attitude adjustment vector based on rotation and translation operators, calculating the deformation residual between geometric feature parameters and pose true coordinates; when the deformation residual exceeds a preset safety threshold, outputting a warning signal for the controlled object and triggering a logical suspension action in the installation process.

[0041] Preferably, the topology consistency arbitration logic further includes: determining the proportion of effective data points in the multi-source spatial feedback signal stream to determine the signal sampling quality; when the signal sampling quality is less than 60%, switching the supervision mode from the global image mode to the key target alignment mode, and generating a correction adjustment signal using the local topology information of the preset physical target to achieve precision control.

[0042] Preferably, the multi-source spatial feedback signal stream in step 102 is real-time point cloud data generated by non-contact detection of the controlled object using a high-frequency laser scanning device; the extraction process of geometric feature parameters includes: denoising the real-time point cloud data, and locking the geometric centroid and axis vector information of the controlled object based on the curvature change point identification algorithm.

[0043] Preferably, the correction adjustment signal is used to drive the end effector unit to generate physical compensation displacement in three dimensions, so as to reduce the real-time pose deviation of the controlled object relative to the topological constraint reference model; the end effector unit includes a multi-axis drive system composed of a hydraulic drive mechanism or a servo module, and the multi-axis drive system completes pose fine adjustment in response to the correction adjustment signal.

[0044] Preferably, when extracting process topology logic nodes in step 101, a topology connection matrix between each process topology logic node is established simultaneously; the topology connection matrix defines the physical assembly constraint relationship of different interface components inside the controlled object; the topology consistency arbitration logic performs logical verification on the pose evolution process of geometric feature parameters based on the topology connection matrix.

[0045] Preferably, in step 103, when establishing the normalized topological deviation field, a nonlinear least squares fitting algorithm is used to iteratively optimize the geometric feature parameters to minimize the sum of squared Euclidean distances between the geometric feature parameters and the true pose coordinates; the normalization process includes mapping the dynamic pose deviation of the controlled object to a preset zero-bias reference space to eliminate system background noise interference.

[0046] Preferably, the precision control method is accomplished through a monitoring program deployed in an industrial control terminal; the monitoring program outputs dynamic correction pulses by comparing the transient installation trajectory of the controlled object with the theoretical path preset by the topological constraint reference model in real time; the industrial control terminal has a real-time communication interface for transmitting trajectory correction instructions to the automated drive module of the controlled object.

[0047] An industrial equipment installation accuracy control system based on BIM and laser fusion positioning includes:

[0048] The topology reference extraction module is used to extract process topology logic nodes that represent key assembly positions of the controlled object from the topology constraint reference model, and to establish the pose true coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor.

[0049] The feedback signal acquisition module is used to acquire multi-source spatial feedback signal streams for the controlled object in real time and extract geometric feature parameters from the multi-source spatial feedback signal streams.

[0050] The topology arbitration calculation module is used to perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and to identify and suppress discrete pose vectors in geometric feature parameters through topology consistency arbitration logic to generate a normalized topology deviation field.

[0051] The correction and adjustment output module is used to extract rotation and translation operators based on the normalized topological deviation field to generate trajectory correction instructions, and send the correction and adjustment signal containing installation displacement compensation information to the end adjustment unit to drive the controlled object to complete the pose control.

[0052] Example 1: In an industrial scenario involving the installation of a semiconductor lithography machine stage base weighing 2000kg, the large base experiences a 0.05mm non-rigid micro-bending due to stress at the support points. This causes the computer monitoring system to fail to decouple global position deviation from local deformation during point-to-point geometric matching, introducing discrete noise containing structural stress characteristics into the control feedback loop. This induces overshoot in the drive mechanism. A computer-monitored control logic based on Building Information Model (BIM) topology association is employed to extract process topology logic nodes representing key assembly positions of the controlled object from the topology constraint reference model. Simultaneously, the true pose coordinates of each process topology logic node in three-dimensional space and the environmental thermal expansion coefficient tensor used to compensate for thermal instability are established. Non-contact detection of the controlled object is performed using a high-frequency laser scanning device to obtain multi-source spatial feedback signal flow, and geometric feature parameters covering edge contours, feature apertures, and surface curvature information are extracted from it. The geometric feature parameters are mapped to the process topology logic nodes in three dimensions in a spatiotemporal manner, and the weighted deviation energy function between physical points and logic nodes is calculated. ;in, The weighted bias energy function, For node indexing, The total number of nodes. For the first Weight parameters of each process topology logical node For the first One geometric feature parameter, For rotation operators, For the first The true pose coordinates of each process topology logic node. As a translation operator, in the process of minimizing the weighted deviation energy function, this invention introduces spatial cooperative constraints between nodes through a topological connectivity matrix, transforming traditional point-to-point geometric registration into morphological evolution analysis based on topological stiffness distribution. The weight parameters... It is not a static constant, but a dynamic variable characterizing the rigidity contribution of nodes in global pose transformation; when the controlled object undergoes non-rigid micro-bending, the offset direction of the nodes affected by deformation will deviate from the pre-set normal consistency of the topology. At this time, the computing system identifies local abnormal gradients in the deviation vector field and utilizes the sensitivity of the energy function to non-consistent offsets to include local distortion terms containing structural stress characteristics. By separating the global rotation operator R and translation operator T, this decoupling mechanism essentially uses the global logical redundancy of the topology to mask local non-rigid defects, so that the convergence direction of the energy function points only to the center of gravity pose transformation of the physical entity, thereby achieving technical decoupling between the global pose and the surface deformation at the algorithm level.

[0053] The system utilizes topology consistency arbitration logic integrated into a computer monitoring system to monitor the deviation trend between physical sampling points and adjacent logical nodes. When the cross-correlation coefficient between the deviation vector and the deviation trend of neighboring nodes is lower than a preset logical threshold, the system automatically reduces the weight parameter of that node. The cross-correlation coefficient of the deviation trend is obtained by calculating the cosine similarity between the pose deviation vector of the current node and the mean deviation vector of the neighboring node set defined by the topological connectivity matrix. The system extracts the relationship between the current node and its neighbors. The spatial offset gradient of each logical node on the three-dimensional coordinate axis is transformed into a normalized trend feature sequence. The cross-correlation coefficient reflects whether the sampling point follows the overall motion trend of the controlled object. If the coefficient is lower than the logical threshold, the point is determined to contain non-random local deformation noise or occlusion interference. This quantization rule provides a definite mathematical basis for topology consistency arbitration, enabling the system to accurately identify discrete pose vectors through scalarization comparison. This utilizes the overall logical integrity of the topology to shield against non-rigid deformation interference caused by gravity or uneven support. The computer monitoring system minimizes the weighted bias energy function. To identify global pose deviations, and combine the environmental thermal expansion coefficient tensor to correct the reference drift caused by environmental temperature difference, and remove the deformation residual terms caused by local stress. To construct a normalized topological deviation field.

[0054] The rotation operator is determined by nonlinear solution based on the topological deviation field. and translation operators To generate trajectory correction commands, read the spatial topological geometric parameters of each support point in the multi-axis drive system, establish a nonlinear Jacobian inverse matrix, call the Jacobian inverse matrix, and apply the rotation operator in the three-dimensional Cartesian coordinate system. With translation operator The decoupled calculation is converted into linear extension stroke and rotation angle commands for independent drive mechanisms. When handling non-rigid equipment installations, this invention employs a local quasi-rigid body iteration strategy to accommodate the linear control logic of the Jacobian inverse matrix. In specific implementation, although the controlled object exhibits minute deformation globally, within a single 300ms monitoring cycle, the system treats the current physical deformation state as a quasi-static constraint, controlling the deformation residual... The compensation is applied to the instantaneous pose of the topological nodes, enabling the controlled object to satisfy the rigid body dynamics assumptions with minimal displacement increments. The Jacobian inverse matrix is ​​only responsible for mapping the geometric pose deviation in this instantaneous state to the actuator. With multiple rounds of measurement-compensation-drive digital closed loop, the logical inaccuracies caused by deformation are eliminated in real time through dynamic reconstruction of weights in the subsequent feedback loop. This control architecture allows the system to utilize a mature rigid body kinematics model to gradually approach the precise positioning target of non-rigid objects through high-frequency micro-step iterations, thereby achieving high-precision closed-loop convergence, generating a discrete control pulse sequence that can be recognized by the underlying hardware, and sending the correction adjustment signal containing installation displacement compensation information to the end adjustment unit. This drives a multi-axis drive system composed of hydraulic drive mechanisms or servo modules to generate physical compensation displacement in three dimensions, enabling the controlled object to complete precise pose control within a 300ms monitoring period. Ultimately, the positioning accuracy of large industrial equipment under complex stress conditions is maintained within 0.01mm.

[0055] Example 2: In simulating industrial precision installation conditions using a physical experimental platform, the platform includes a high-frequency laser scanning device with a ranging accuracy of 0.02mm and a sampling frequency of 50Hz, and a multi-axis drive system composed of servo modules. The experiment uses the high-frequency laser scanning device to acquire multi-source spatial feedback signal streams of the controlled object under dust interference. The sampling period is set based on a technical trade-off between the transient displacement components of the controlled object's trajectory and the convergence time of the algorithm's nonlinear solution. When the controlled object's movement speed is between 5mm / s and 10mm / s, a monitoring period of 300ms is selected to provide geometric feature parameters with sufficient spatial discriminative power. The control loop incorporates a feedforward prediction mechanism to read the real-time drive rate of the controlled object and the acceleration commands issued by the drive mechanism. A state space matrix containing position and velocity parameters is established. The expected displacement increment at the end of the current monitoring cycle is iteratively extrapolated using a Kalman filter algorithm as a pre-offset term and dynamically superimposed onto the three-dimensional coordinate system of the process topology logic node. Under a baseline condition with 85% signal sampling quality, the computer monitoring system extracts geometric feature parameters. and with process topology logic nodes Perform three-dimensional spatiotemporal mapping according to the formula Calculate the weighted bias energy function; where, The weighted bias energy function, For node indexing, The total number of nodes. For the first Weight parameters of each process topology logical node For the first One geometric feature parameter, For rotation operators, For the first The true pose coordinates of each process topology logic node. For translation operators; since the initial input signal quality is stable, the system will assign weight parameters to each process topology logic node. All values ​​remain at 1.00. The calculated global pose deviation is then processed by the rotation operator. With translation operator The controllable object is converted into a trajectory correction command and positioned under the action of the multi-axis drive system. The measured positioning accuracy is 0.009mm.

[0056] Under conditions where signal sampling quality drops to 70% due to increased dust concentration, the control group, employing point-to-point geometric matching logic, experiences a cumulative positioning error of 0.218 mm. In contrast, the present invention's sample group initiates topological consistency arbitration logic to monitor the deviation trend between physical sampling points and adjacent logical nodes. When the cross-correlation coefficient between the sampling point deviation vector in the dust-covered area and the deviation trend of neighboring nodes falls below the logic threshold of 0.45, the system determines that the sampling point is a discrete pose vector and adjusts its corresponding weight parameters. To 0.15, the overall logical integrity of the topology is used to shield against non-rigid deformation interference, and the weighted bias energy function is minimized. Re-extract the rotation operator and translation operators The calculated deformation residuals It is 0.012mm.

[0057] When the signal sampling quality drops to 55% in an edge condition, the system identifies that the proportion of effective data points is less than a preset threshold of 60%. The supervision mode switches from global imaging mode to key target alignment mode. The system uses the local topology information of the preset physical target to generate a correction adjustment signal. The experimental data shows that the control group experiences driving oscillations and the positioning accuracy deteriorates to more than 0.530 mm. However, the sample group of this invention generates a correction adjustment signal containing installation displacement compensation information through the key target alignment mode, which drives the end adjustment unit to generate physical compensation displacement in three dimensions. The positioning accuracy of the controlled object is maintained at 0.010 mm. Based on the above gradient test data, this invention eliminates the influence of discrete noise on the supervision control feedback loop in a dust-covered environment through computer supervision and control logic of building information model topology association. This enables dynamic synchronization between the control benchmark and the physical entity state, and maintains the positioning accuracy of large industrial equipment within 0.01 mm under complex stress conditions.

[0058] Example 3: In an industrial assembly workshop containing multiple metal reflective surfaces, when a high-frequency laser scanning device probes a controlled object, it generates a large number of specular reflection noise points due to the specular reflection from the metal surface. These specular reflection noise points overlap with the physical edges of the controlled object in three-dimensional space, leading to local geometric feature parameters... The spatial distribution density exhibits non-random clustering, making it difficult for computer monitoring systems to identify effective assembly features using a single distance metric when performing three-dimensional spatiotemporal mapping.

[0059] To address signal distortion caused by specular reflection, the computer monitoring system employs topology consistency arbitration logic based on local feature intensity evaluation at the process topology logic nodes. In the topologically constrained reference model, the curvature change rate of the key assembly position of the controlled object is considered. Determine the spatial sampling step size Spatial sampling step size With rate of change of curvature To satisfy the inverse proportional function relationship, the distribution density of process topology logic nodes in high curvature feature regions is maintained at 5 nodes / cm. 2 The multi-source spatial feedback signal stream of the controlled object's physical entity is acquired using a high-frequency laser scanning device. Local geometric feature parameters are extracted from the multi-source spatial feedback signal stream using a three-dimensional spatial sliding window. The local variation coefficient of the sampling pose vector of each point cloud within the three-dimensional spatial sliding window is then calculated. To characterize the discreteness of the signal, and using the physical assembly constraints defined by the topological connection matrix, a real-time topological feature vector is generated by calculating the pose deviation gradient between adjacent physical sampling points.

[0060] Topological consistency arbitration logic is based on local variation coefficients. The statistical distribution pattern automatically determines the logical threshold of the cross-correlation coefficient. Logical threshold The calculation process is as follows: Calculate the mean of the local coefficient of variation within the preset time window. and standard deviation The logical threshold is determined as ;in, For logical thresholds, This represents the mean of the local coefficients of variation. The standard deviation of the local coefficient of variation. The sensitivity coefficient, calibrated based on environmental reflectivity, is set to 1.2 under this operating condition. The computer monitoring system performs cross-correlation calculations between the real-time topological feature vector and the true topological vector in the topological constraint reference model. When the obtained cross-correlation coefficient is lower than the logical threshold... When the corresponding geometric feature parameter is determined to be a discrete pose vector generated by mirror reflection, the weight parameter of the node is linearly reduced according to the deviation. Weight parameters The calculation method is as follows: ;in, For the first Weight parameters of each process topology logical node For logical thresholds, The local coefficient of variation for the corresponding sampling point; through the weighted bias energy function To reduce the energy proportion of anomalous point clouds and suppress noise interference, in an environment containing 30% specular reflection noise points, the computer supervision system identifies the skeleton features representing the pose of physical entities, removes feature artifacts caused by specular reflection, and completes the rotation operator. With translation operator The high-precision calculation generates trajectory correction commands that drive the end-effector adjustment unit to produce sub-millimeter-level pose compensation actions. In industrial sites with multipath interference, the positioning deviation of the equipment installation is controlled within 0.012mm, achieving digital closed-loop synchronization between the control reference and the physical entity state.

[0061] Example 4: In the on-site deployment scenario of an ultra-large flexible assembly line, due to the randomness of the installation position of the high-frequency laser scanning device, there is a spatial topological deviation between its own measurement coordinate system and the global process coordinate system defined by the building information model. The computer monitoring system calibrates the initial pose transformation matrix by detecting a set of fixed control points preset around the installation position of the controlled object. By extracting the true pose coordinates of the fixed control points in three-dimensional space and comparing them with the sampled pose vector returned by the high-frequency laser scanning device, the rotation increment and translation increment along each coordinate axis are calculated to achieve the initial alignment of the coordinate system. The system then uses the consistency of the local curvature distribution at key intersections based on the laser point cloud features to iteratively optimize and limit the initial coincidence error of the spatial reference to within 0.005mm, supporting subsequent topological mapping.

[0062] When the controlled object faces an ambient temperature fluctuation from 20℃ to 28℃, the thermophysical properties of the material cause scaling with temperature changes, resulting in deviations of the process topology logic nodes from the original model baseline. The computer monitoring system constructs an environmental thermal expansion coefficient tensor based on a preset material property constant table and real-time temperature gradient fields fed back by temperature sensors distributed at key points of the controlled object. The system fits a continuous temperature field by collecting discrete temperature values ​​from a thermocouple array. For extracted process topology logic nodes, it calculates the thermal deformation compensation vector along each orthogonal physical axis. The specific calculation formula is as follows: ,in, To specify the linear expansion component along the physical axis, To obtain the inherent linear thermal expansion coefficient of the main material of the controlled object from a table, The distance is limited to the initial spatial distance from the process topology logic node to the thermally rigid constraint reference point. and The real-time continuous temperature acquisition of the region and the reference temperature determined during the system calibration phase are respectively characterized, and the linear expansion components of each specified physical axis are calculated based on the above. , , This invention constructs an environmental thermal expansion coefficient tensor by mapping it to corresponding spatial transformation components. Mathematically, this tensor is a 3×3 diagonal matrix, with its main diagonal elements representing the thermal strain components of the controlled object along the three orthogonal principal stress axes. Using the continuous temperature field gradient obtained through a thermocouple array, the system can calculate the anisotropic deformation characteristics in the tensor in real time and accordingly adjust the process topology logic nodes. The true value of the three-dimensional spatial pose is compensated and corrected. This transformation process ensures that the overall environmental temperature variable can be converted into coordinate fine-tuning quantities for each node of the topological constraint reference model through a tensor mapping mechanism, eliminating the logical gap between scalar calculation and tensor definition. By calculating the linear expansion components of the material along the orthogonal physical axes, thermal deformation compensation vectors are generated for each process topological logic node. These thermal deformation compensation vectors are then superimposed on the true pose coordinates in real time, completing the dynamic correction of the topological constraint reference model. This allows the system to extract deformation residuals under temperature gradient disturbances. Characterizes the structural deformation components induced by physical installation actions.

[0063] Example 5: In the initialization phase involving the deployment of a flexible assembly line for aero-engines, due to structural vibrations in the plant and random deviations in the installation posture of the high-frequency laser scanning device, the measurement reference is physically offset. The computer monitoring system initiates the reference field calibration process, arranging eight fixed control points around the installation position of the controlled object. The high-frequency laser scanning device performs ten repeated physical samplings on the fixed control point group to construct an initial posture statistical library. The system determines the sensitivity coefficient based on the variance envelope of the sampling point coordinates in three-dimensional space. The baseline value is determined, and the detection deviation caused by hardware thermal disturbance is characterized by calculating the cumulative root mean square error of the sampling point displacement within a preset time window. Based on this, a three-dimensional spatial sliding window is used to perform real-time signal-to-noise ratio calculation on the multi-source spatial feedback signal stream. The control program calculates the effective sampling weight corresponding to each process topology logic node through a nonlinear mapping function. The calculation formula is as follows: ,in, These are dimensionless weight parameters based on the topological logic nodes of each process, and their values ​​are constrained to a closed interval between zero and one. To extract and calculate the real-time signal-to-noise ratio of local sampling points using a three-dimensional sliding window, A preset signal-to-noise ratio threshold is used to define the boundary of signal quality degradation. To determine the weight attenuation slope gain coefficient and fix it as a positive number, the real-time signal-to-noise ratio of local sampling points is... The signal strength is determined based on the statistical characteristics of spatial point cloud density. Within a three-dimensional sliding window, the system counts the number of valid points falling within the preset tolerance radius of the process topology logic nodes as the signal strength component, and considers the number of isolated points that deviate from the topology skeleton and exhibit a scattered distribution within the window as the background noise component. The signal-to-noise ratio is... This is the ratio of the signal component to the noise component, determined by a pre-set signal-to-noise ratio threshold. The system can automatically identify the degree of data contamination caused by drastic changes in lighting or dust scattering. It then dynamically adjusts the energy weights of each topology node participating in pose calculation through the aforementioned logic function, ensuring the purity of the feedback signal source and adjusting the initial weight parameters of each process topology logic node. Set to a value positively correlated with the signal-to-noise ratio, where, For the first The weight parameters of each process topology logical node are used to establish a mapping relationship between the global process coordinate system and the local measurement coordinate system with an overlap error of less than 0.005mm before equipment installation.

[0064] Once the system enters the online precision pose control process, the topology consistency arbitration logic uses the initial pose statistics library generated during the calibration phase to monitor geometric feature parameters. When the deviation trend is detected, and physical deformation of the controlled object caused by the support stress is detected, the system automatically calls the environmental thermal expansion coefficient tensor to the process topology logic node. The pose true coordinates are corrected in real time, and the second-order rate of change of the cross-correlation coefficient is used as the trigger condition for switching the monitoring mode. If the rate of decrease of the cross-correlation coefficient exceeds the preset dynamic slope threshold for five consecutive sampling periods, the system automatically limits the energy weight of global features in the deviation calculation and starts the key target alignment mode. By extracting the local topology information of the physical target, a correction and adjustment signal containing installation displacement compensation information is generated, which drives the end adjustment unit composed of hydraulic drive mechanism or servo module to generate physical displacement components in three dimensions until the deformation residual of the controlled object is corrected. By converging to a safe range within 0.012mm, large industrial equipment can achieve precise position locking under varying stress conditions.

[0065] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning, characterized in that, Includes the following steps: Step 101: Extract the process topology logic nodes that represent the key assembly positions of the controlled object in the topology constraint reference model, and establish the true pose coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor used to compensate for thermal instability. Step 102: Real-time acquisition of multi-source spatial feedback signal streams for the physical entity of the controlled object, and extraction of geometric feature parameters covering edge contours, feature apertures and surface curvature information from the multi-source spatial feedback signal streams; Step 103: Perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and identify and suppress discrete pose vectors in geometric feature parameters that exceed the preset deviation threshold due to dust occlusion or multiple reflections through topology consistency arbitration logic integrated into the computer supervision system, thereby constructing a normalized topology deviation field. Step 104: Based on the normalized topological deviation field, the rotation operator and translation operator are extracted through nonlinear calculation to generate trajectory correction instructions, and the correction adjustment signal containing installation displacement compensation information is sent to the end adjustment unit. The controlled object is driven to complete precise pose control through digital closed-loop logic supervision. The topology consistency arbitration logic in step 103 includes: calculating the deviation vector of each geometric feature parameter relative to the process topology logic node; when the cross-correlation coefficient between the deviation vector and the deviation trend of neighboring nodes is lower than a preset logic threshold, automatically reducing the weight parameter of the node, and using the logic integrity of the remaining nodes to maintain the output stability of the topology deviation field.

2. The industrial equipment installation accuracy control method based on BIM and laser fusion positioning according to claim 1, characterized in that, The trajectory correction instructions generated in step 104 also include: generating a six-degree-of-freedom attitude adjustment vector based on rotation and translation operators, calculating the deformation residual between geometric feature parameters and pose true coordinates; and when the deformation residual exceeds a preset safety threshold, outputting a warning signal for the controlled object and triggering a logical suspension action in the installation process.

3. The method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning according to claim 2, characterized in that, The topology consistency arbitration logic also includes: determining the proportion of effective data points in the multi-source spatial feedback signal stream to determine the signal sampling quality; when the signal sampling quality is below 60%, switching the supervision mode from global image mode to key target alignment mode, and using the local topology information of the preset physical target to generate a correction adjustment signal to achieve precision control.

4. The method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning according to claim 1, characterized in that, In step 102, the multi-source spatial feedback signal stream is real-time point cloud data generated by non-contact detection of the controlled object using a high-frequency laser scanning device; the extraction process of geometric feature parameters includes: denoising the real-time point cloud data, and locking the geometric centroid and axis vector information of the controlled object based on the curvature change point identification algorithm.

5. The industrial equipment installation accuracy control method based on BIM and laser fusion positioning according to claim 1, characterized in that, The correction adjustment signal is used to drive the end effector unit to generate physical compensation displacement in three dimensions, so as to reduce the real-time pose deviation of the controlled object relative to the topological constraint reference model. The end effector unit includes a multi-axis drive system composed of a hydraulic drive mechanism or a servo module. The multi-axis drive system responds to the correction adjustment signal to complete the pose fine adjustment.

6. The method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning according to claim 1, characterized in that, When extracting process topology logic nodes in step 101, a topology connection matrix between each process topology logic node is established simultaneously; the topology connection matrix defines the physical assembly constraint relationship between different interface components inside the controlled object. The topology consistency arbitration logic performs logical verification on the pose evolution process of geometric feature parameters based on the topology connection matrix.

7. The industrial equipment installation accuracy control method based on BIM and laser fusion positioning according to claim 1, characterized in that, In step 103, when establishing the normalized topological deviation field, a nonlinear least squares fitting algorithm is used to iteratively optimize the geometric feature parameters in order to minimize the sum of squared Euclidean distances between the geometric feature parameters and the true pose coordinates. The normalization process includes mapping the dynamic pose deviation of the controlled object to a preset zero-bias reference space to eliminate system background noise interference.

8. The method for controlling the installation accuracy of industrial equipment based on BIM and laser fusion positioning according to claim 1, characterized in that, The precision control method is accomplished through a monitoring program deployed in an industrial control terminal. The monitoring program outputs dynamic correction pulses by comparing the transient installation trajectory of the controlled object with the theoretical path preset by the topological constraint reference model in real time. The industrial control terminal has a real-time communication interface for transmitting trajectory correction commands to the automated drive module of the controlled object.

9. An industrial equipment installation accuracy control system based on BIM and laser fusion positioning, used to implement the industrial equipment installation accuracy control method based on BIM and laser fusion positioning as described in claim 1, characterized in that, include: The topology reference extraction module is used to extract process topology logic nodes that represent key assembly positions of the controlled object from the topology constraint reference model, and to establish the pose true coordinates of each process topology logic node in three-dimensional space as well as the environmental thermal expansion coefficient tensor. The feedback signal acquisition module is used to acquire multi-source spatial feedback signal streams for the controlled object in real time and extract geometric feature parameters from the multi-source spatial feedback signal streams. The topology arbitration calculation module is used to perform three-dimensional spatiotemporal mapping between geometric feature parameters and process topology logic nodes, and to identify and suppress discrete pose vectors in the geometric feature parameters through topology consistency arbitration logic to generate a normalized topology deviation field. The topology consistency arbitration logic includes: calculating the deviation vector of each geometric feature parameter relative to the process topology logic node; when the cross-correlation coefficient between the deviation vector and the deviation trend of neighboring nodes is lower than a preset logic threshold, the weight parameter of that node is automatically reduced, and the output stability of the topology deviation field is maintained by utilizing the logical integrity of the remaining nodes. The correction and adjustment output module is used to extract rotation and translation operators based on the normalized topological deviation field to generate trajectory correction instructions, and send the correction and adjustment signal containing installation displacement compensation information to the end adjustment unit to drive the controlled object to complete the pose control.

Images