A wind farm self-adaptive group tower operation anti-collision method

Abstract

The application discloses a wind field self-adaptive group tower operation anti-collision method, which specifically comprises the following steps: obtaining static parameters and installation positions of each tower crane and establishing a unified space coordinate system; collecting a rotary angle, an amplitude, a hook height, speed parameters and a wind speed and direction in operation according to a sampling period; calculating three-dimensional coordinates of a crane boom end, a hook and a rear bridge end and constructing a three-dimensional operation envelope; obtaining a dynamic safety distance threshold value according to a wind field correction base safety distance and performing directional correction according to a relationship between a wind direction and a rotary direction; filtering an envelope body sequence, establishing a relative motion model containing wind field disturbance, predicting a prediction envelope body in a prediction time window and performing lag compensation; calculating a minimum safety gap of the prediction envelope body and comparing the minimum safety gap with the threshold value to output an early warning or a control instruction. The method can reduce false alarms and missed alarms, improve determination stability and field applicability.

Description

Technical Field

[0001] This invention relates to the field of intelligent safety monitoring technology for tower cranes, and in particular to a wind-adaptive method for preventing collisions during multi-tower operations. Background Technology

[0002] On construction sites, tower cranes are often deployed simultaneously for coordinated lifting operations. Due to space constraints, the slewing radius, boom coverage area, and hook operating space of each tower crane can easily overlap. Especially during peak operating periods, the overlapping boom slewing, luffing, and lifting actions of multiple tower cranes pose a risk of them approaching each other or even colliding. To reduce the probability of accidents, collision warning devices or distance-based limit control schemes are typically used on-site to monitor the relative positions of critical parts of the tower cranes and issue alarms or control commands.

[0003] Current common collision avoidance technologies primarily rely on spatial distance calculations, determining whether to trigger a warning or brake based on a preset fixed safety distance threshold. This approach often suffers from insufficient adaptability in engineering applications. When environmental wind speed and direction change, the boom and load may shift or sway, causing deviations between the actual operating space of critical components and the ideal geometric model. However, the fixed threshold does not adjust with changing operating conditions, easily leading to situations where the threshold is too high, resulting in frequent false alarms, or too low, preventing timely risk detection.

[0004] Furthermore, some existing technologies primarily focus on static or instantaneous distance relationships when assessing risks, lacking a comprehensive consideration of the tower crane's own motion state and environmental factors. For example, changes in the tower crane's slewing, luffing, and hoisting speeds can alter the approach speed and direction of critical components; wind speed and direction can also cumulatively affect the swing amplitude of the load and the relative motion trend of critical components. If relying solely on a single distance indicator or simple threshold comparison, the early warning results often struggle to simultaneously ensure accuracy and stability, especially in complex scenarios involving multiple tower cranes working together.

[0005] Meanwhile, tower crane operations have significant dynamic and time-sensitive requirements. During high-speed rotation or continuous movement of the tower crane boom, there is an inevitable delay in sensor acquisition, communication synchronization, and computational processing. If the existing system lacks a mechanism to handle this delay, it is prone to judgment lag, resulting in situations where the risk is approaching but the system alerts too late, thus affecting the reliability of collision avoidance control.

[0006] From a construction organization perspective, some existing solutions tend to adopt a rigid shutdown strategy after a risk is triggered, lacking a more detailed tiered early warning and gradual control mechanism, as well as a controllable manual intervention method under the premise of ensuring safety constraints. As a result, frequent system alarms or braking may cause unnecessary downtime, affecting hoisting efficiency and continuous operation rhythm, and making on-site management more reliant on experience-based handling. Summary of the Invention

[0007] To address the problems of existing anti-collision technologies in collaborative tower crane operations, which generally rely on fixed safety distance thresholds and thus lack adaptability to changes in wind speed and direction, rely solely on distance judgment without considering the coupling analysis of tower crane motion status and environmental factors leading to low early warning accuracy, suffer from judgment lag due to data acquisition and processing delays in dynamic scenarios such as high-speed rotation, and have rigid control strategies after triggering affecting continuous construction, this invention provides a wind-adaptive anti-collision method and system for tower crane operations. By collecting and fusing static parameters of tower cranes, real-time operating status data, and wind speed and direction information, a unified spatial model of key parts of the tower cranes is constructed. Risk assessment is performed based on dynamic safety thresholds corrected by the wind field and short-term trajectory prediction. Combined with trajectory smoothing processing with lag compensation, hierarchical early warning, and progressive control mechanisms, safe and efficient anti-collision for tower cranes is achieved.

[0008] To achieve the above-mentioned objectives, this invention provides a wind field-adaptive method for preventing collisions in multi-tower operations, the method comprising the following steps: Step S1: Obtain the static parameters and installation position information of each tower crane in the multi-tower operation scenario. The static parameters include tower height parameters, boom geometry parameters, rear axle geometry parameters, and foundation safety distance. Establish a unified spatial coordinate system based on the installation position information. Step S2: During the operation, acquire the real-time operating status parameters and wind field parameters of each tower crane at a preset sampling period; the real-time operating status parameters include slewing angle, amplitude, hook height, slewing speed, luffing speed and hoisting speed, and the wind field parameters include wind speed and wind direction; Step S3: Based on the static parameters, the real-time operating status parameters, and the unified spatial coordinate system, calculate the three-dimensional coordinates of the boom end, hook, and rear axle end of each tower crane, and construct the three-dimensional working envelope of each tower crane based on the three-dimensional coordinates; Step S4: Correct the basic safety distance based on the wind field parameters to obtain a dynamic safety distance threshold, and make directional corrections to the dynamic safety distance threshold based on the relationship between wind direction and rotation direction; Step S5: Filter the time series of the three-dimensional working envelope, and establish a relative motion model based on the slewing speed, the amplitude change speed, the lifting speed and the wind field parameters; Step S6: Based on the relative motion model, perform short-time trajectory prediction on the three-dimensional working envelope of different tower cranes to obtain the predicted envelope within the prediction time window, and perform hysteresis compensation on the predicted envelope when the preset conditions are met. Step S7: Calculate the minimum safe clearance between different tower cranes in the predicted envelope within the predicted time window, and compare the minimum safe clearance with the dynamic safe distance threshold to obtain the risk assessment result; output the early warning command or control command according to the risk assessment result.

[0009] Preferably, in step S1, the unified spatial coordinate system is a global coordinate system established with the fixed reference point of the construction site as the coordinate origin and the horizontal plane as the reference plane, and the installation position information of each tower crane is represented as the coordinates of the center point of the tower body under the global coordinate system. The static parameters include the maximum working radius of the boom and the maximum overhang of the rear axle. These static parameters are entered into the tower crane model parameter table and stored in the storage module during system deployment.

[0010] Preferably, in step S2, both the real-time operating status parameters and the wind field parameters are timestamped. The rotational speed is calculated from the difference in rotational angle between adjacent sampling periods; The amplitude variation rate is calculated from the amplitude difference between adjacent sampling periods; The lifting speed is calculated from the difference in hook height between adjacent sampling periods; The wind field parameters are collected by wind speed and direction sensors installed on the tower crane structure.

[0011] Preferably, in step S3, before calculating the three-dimensional coordinates of the end of the boom, the hook, and the end of the rear axle, the collected data of different tower cranes are time-aligned according to the timestamp carried in step S2. When calculating the three-dimensional coordinates of the boom end, hook and rear axle end of each tower crane, the center point of the tower body is used as the reference point for coordinate calculation. The coordinates of the center point of the tower body in the unified spatial coordinate system are superimposed with the horizontal displacement determined by the rotation angle and the amplitude and the vertical displacement determined by the hook height to obtain the corresponding three-dimensional coordinates. When constructing the three-dimensional working envelope, the boom sweep envelope and the rear axle sweep envelope are determined according to the boom geometry parameters and the rear axle geometry parameters, and the hook spreader envelope is determined according to the three-dimensional coordinates of the hook and the preset outward expansion. Then, the boom sweep envelope, the rear axle sweep envelope and the hook spreader envelope are merged to construct the three-dimensional working envelope.

[0012] Preferably, in step S4, the dynamic safety distance threshold is obtained by adding the basic safety distance and the wind speed correction amount. The wind speed correction amount increases with the increase of wind speed and is segmented according to the preset wind speed level. The wind speed used to determine the wind speed correction amount is a weighted combination of instantaneous wind speed and the average wind speed of the preset time window.

[0013] Preferably, in step S4, the rotation direction is determined by the direction of change of the rotation angle within adjacent sampling periods. When the rotation angle increases, it is determined as the first rotation direction, and when the rotation angle decreases, it is determined as the second rotation direction. The directional correction includes determining the directional correction amount based on the angle between the wind direction and the rotation direction, and dividing the directional correction amount into three correction states: tailwind correction, crosswind correction, and headwind correction, according to the range of the angle. Tailwind correction corresponds to increasing the dynamic safety distance threshold, headwind correction corresponds to decreasing the dynamic safety distance threshold, and crosswind correction corresponds to keeping the dynamic safety distance threshold unchanged.

[0014] Preferably, in step S5, filtering the time series of the three-dimensional working envelope includes performing sliding time window filtering on the coordinate sequence representing the position of the end of the crane boom, the hook and the end of the rear axle, and performing amplitude limiting processing on the filtered coordinate sequence to suppress short-term abrupt changes caused by the swing of the hook and lifting device, thereby obtaining a smooth envelope sequence for establishing the relative motion model.

[0015] Preferably, in step S5, the relative motion model determines the instantaneous motion velocity components of the three-dimensional working envelope based on the slewing speed, luffing speed, and hoisting speed, and introduces the wind speed and wind direction corresponding to the wind field parameters as external disturbance terms into the instantaneous motion velocity components, thereby obtaining a relative motion model for short-term trajectory prediction between different tower cranes.

[0016] Preferably, in step S6, the short-time trajectory prediction includes, based on the relative motion model established in step S5, performing discrete time-time recursive prediction of the representative points of the three-dimensional operation envelope within the prediction time window, and generating the prediction envelope based on the predicted position sequence of the representative points; The prediction envelope includes combining the predicted position sequence of the representative points with the geometric outline parameters of the three-dimensional operation envelope, and applying an expansion amount to the predicted position sequence of the representative points to obtain a prediction envelope that covers the spatial occupancy range within the prediction time window. The expansion amount is used to characterize the spatial outline expansion caused by wind field disturbance and hook and hoist swing. The hysteresis compensation includes fitting the projection trajectory of the predicted envelope on a preset horizontal plane to obtain a fitted trajectory when the rotation speed meets the preset conditions, and correcting the position of the predicted envelope based on the fitted trajectory to compensate for the prediction deviation caused by the hysteresis of the rotation drive response or the sampling delay.

[0017] Preferably, in step S7, the minimum safety gap is the minimum distance between the prediction envelopes of different tower cranes within the prediction time window. The minimum safety gap is obtained by calculating the envelope distance for each discrete prediction time within the prediction time window and taking the minimum value. When calculating the minimum safety gap, first apply uncertainty boundary expansion to the predicted envelope to obtain a conservative predicted envelope, and then calculate the minimum distance between the conservative predicted envelopes of different tower cranes as the minimum safety gap. The risk assessment result includes a warning level and a control level. When the minimum safety gap is less than the warning distance threshold, a warning command is output. When the minimum safety gap is less than the control distance threshold, a control command is output. The warning distance threshold and the control distance threshold are both determined based on the dynamic safety distance threshold.

[0018] The beneficial effects of this invention are as follows: In multi-tower operation scenarios, this invention enables unified modeling and continuous updating of the spatial occupancy range and safety distance criteria for tower cranes, and completes prediction and judgment before risks occur. This reduces false alarms and missed alarms caused by fixed safety distances under different wind conditions and different operational actions, improves the stability and consistency of collision avoidance judgment, and makes early warning and control outputs more in line with real-time requirements on site. Regarding safety distance criteria, by dynamically correcting the basic safety distance based on wind speed and combining it with the relationship between wind direction and rotation direction for directional correction, the safety distance threshold can adaptively adjust with changes in wind conditions, reducing false judgments caused by threshold mismatch when wind speed and direction change. Regarding the risk judgment object, by calculating the three-dimensional coordinates of key parts and constructing a three-dimensional operational envelope in a unified spatial coordinate system, the judgment is no longer limited to single-point distance but covers the boom sweep, rear axle extension, and hook lifting area, reducing risk omissions caused by localized point selection. In terms of early warning, a relative motion model is established based on operating speed parameters and wind field parameters, and a predicted envelope is obtained through short-time trajectory prediction. This gives the system advance capability and allows it to output warnings or control commands before potential collisions. Simultaneously, under conditions such as high slewing, hysteresis compensation is performed through projected trajectory fitting to improve prediction stability and reduce deviations caused by response lag or sampling delay. Regarding reliability, short-time abrupt changes are suppressed by filtering the envelope time series, and an uncertainty boundary extension is introduced into the gap calculation to form a conservative judgment. This ensures that risk assessment maintains a safety margin even with measurement errors, time consistency errors, and wind field fluctuations, reducing the risk of missed alarms under boundary conditions. Attached Figure Description

[0019] Figure 1 This is a flowchart of Embodiment 1 of the present invention. Detailed Implementation

[0020] To clearly illustrate the technical features of this solution, the following detailed implementation method will be used to explain the solution.

[0021] Example 1 See Figure 1 This invention provides a wind field adaptive method for collision avoidance in multi-tower operations, which includes the following steps: Step S1: Obtain the static parameters and installation position information of each tower crane in the multi-tower operation scenario. The static parameters include tower height parameters, boom geometry parameters, rear axle geometry parameters, and foundation safety distance. Establish a unified spatial coordinate system based on the installation position information. The unified spatial coordinate system is a global coordinate system established with a fixed reference point of the construction site as the origin and the horizontal plane as the reference plane. The installation position information of each tower crane is represented as the coordinates of the tower center point under the global coordinate system. The static parameters include the maximum working radius of the boom and the maximum overhang of the rear axle. The static parameters are entered from the tower crane model parameter table and stored in the storage module during system deployment.

[0022] Step S2: During the operation, acquire the real-time operating status parameters and wind field parameters of each tower crane at a preset sampling period, as well as the corresponding timestamps; the real-time operating status parameters include slewing angle, amplitude, hook height, slewing speed, luffing speed, and hoisting speed; the slewing speed is calculated from the difference in slewing angle between adjacent sampling periods; the luffing speed is calculated from the difference in amplitude between adjacent sampling periods; the hoisting speed is calculated from the difference in hook height between adjacent sampling periods; the wind field parameters are collected by wind speed and direction sensors installed on the tower crane structure.

[0023] Step S3: Based on static parameters, real-time operating status parameters, and a unified spatial coordinate system, calculate the three-dimensional coordinates of the boom end, hook, and rear axle end of each tower crane, and construct the three-dimensional working envelope of each tower crane based on the three-dimensional coordinates.

[0024] Before calculating the three-dimensional coordinates of the boom end, hook and rear axle end, the collected data of different tower cranes are time-aligned according to the timestamps carried in step S2. When calculating the three-dimensional coordinates of the boom end, hook and rear axle end of each tower crane, the center point of the tower body is used as the reference point for coordinate calculation. The coordinates of the tower body center point in a unified spatial coordinate system are superimposed with the horizontal displacement determined by the slewing angle and amplitude and the vertical displacement determined by the hook height to obtain the corresponding three-dimensional coordinates. When constructing the three-dimensional working envelope, the boom sweep envelope and the rear axle sweep envelope are determined based on the boom geometry parameters and the rear axle geometry parameters. The hook spreader envelope is determined based on the three-dimensional coordinates of the hook and the preset outward expansion. The boom sweep envelope, the rear axle sweep envelope and the hook spreader envelope are then combined to construct the three-dimensional working envelope.

[0025] Step S4: Correct the basic safety distance based on wind field parameters to obtain a dynamic safety distance threshold, and then perform directional correction on the dynamic safety distance threshold based on the relationship between wind direction and rotation direction. This dynamic safety distance threshold is obtained by adding the basic safety distance to the wind speed correction amount. The wind speed correction amount increases with increasing wind speed and is segmented according to preset wind speed levels. The wind speed used to determine the wind speed correction amount is a weighted combination of instantaneous wind speed and the average wind speed over a preset time window. The rotation direction is determined by the direction of change of the rotation angle within adjacent sampling periods. When the rotation angle increases, it is determined as the first rotation direction; when the rotation angle decreases, it is determined as the second rotation direction. The directional correction includes determining the directional correction amount based on the angle between the wind direction and the rotation direction, and dividing the directional correction amount into three correction states based on the angle range: downwind correction, crosswind correction, and headwind correction. Downwind correction corresponds to increasing the dynamic safety distance threshold, headwind correction corresponds to decreasing the dynamic safety distance threshold, and crosswind correction corresponds to keeping the dynamic safety distance threshold unchanged.

[0026] Step S5: Filter the time series of the three-dimensional operation envelope and establish a relative motion model based on the slewing speed, luffing speed, lifting speed and wind field parameters; wherein, filtering the time series of the three-dimensional operation envelope includes performing sliding time window filtering on the coordinate sequence representing the position of the boom end, hook and rear axle end, and performing amplitude limiting processing on the filtered coordinate sequence to suppress short-term abrupt changes caused by the swing of the hook and spreader, thereby obtaining a smooth envelope sequence for establishing the relative motion model.

[0027] The relative motion model determines the instantaneous motion velocity components of the three-dimensional working envelope based on the slewing speed, luffing speed, and hoisting speed. The wind speed and wind direction corresponding to the wind field parameters are introduced as external disturbance terms into the instantaneous motion velocity components, thereby obtaining a relative motion model for short-term trajectory prediction between different tower cranes.

[0028] In one embodiment, after constructing a three-dimensional operational envelope and obtaining a dynamic safety distance threshold, the controller filters the time series of the three-dimensional operational envelope and establishes a relative motion model based on the filtering result to provide a smooth, continuous, and computable input for subsequent short-term trajectory prediction. Because the hook and lifting device in multi-tower operations are subject to wind disturbances and swaying, and sensor sampling and calculation introduce jitter, directly using unprocessed data for prediction can easily lead to significant fluctuations in the prediction results within a short period, resulting in false alarms or frequent changes in control commands. Therefore, the controller first smooths the key boundary position sequence of the envelope using a time window, and then suppresses short-term abrupt changes through a limiting strategy to obtain a stable envelope sequence.

[0029] In one implementation, the controller performs sliding time window filtering on the coordinate sequence of the boom end, hook, and rear axle end in a unified spatial coordinate system. The boom end is located at a timestamp... The three-dimensional coordinates, the filtered coordinates are as follows and Using a length of The average filtering is performed using a sliding time window, expressed as:

[0030] The coordinate sequences of the hook and the rear axle end are obtained using the same method to obtain the corresponding filtered coordinates. To further suppress short-term abrupt changes caused by hook and spreader swing or sampling anomalies, the controller performs amplitude limiting processing on the filtering results, ensuring that the coordinate changes between adjacent sampling times do not exceed a preset amplitude limiting threshold, thereby obtaining a smooth envelope sequence for prediction.

[0031] After obtaining the smooth envelope sequence, a relative motion model is established. The relative motion model describes the motion trend of the envelope within a short prediction time window. Its inputs include the velocity component caused by the tower crane's own motion and the external disturbance component caused by wind field disturbances. For the ... For a tower crane, the slewing speed, luffing speed, and hoisting speed are denoted as follows: and The controller determines the instantaneous velocity component of the hook position based on the slewing speed, luffing speed, and hoisting speed, and adds the wind field disturbance determined by the wind speed and wind direction as an external disturbance term to the instantaneous velocity component, thereby obtaining the motion speed used for prediction.

[0032] In one implementation, the horizontal velocity component of the hook is formed by the combined rotational and luffing motions. The horizontal velocity component caused by the rotational motion is related to the horizontal projection radius of the hook and the rotational speed, while the horizontal velocity component caused by the luffing motion is along the radial direction of the hook. The controller determines the radial and tangential directions based on the rotation angle and maps the rotational and luffing speeds into horizontal velocity components. The vertical velocity component of the hook is determined by the lifting speed. The wind field disturbance term is determined based on wind speed and direction and decomposed into horizontal disturbance components along the coordinate axes of a unified spatial coordinate system. These components are then superimposed on the horizontal velocity component of the hook, enabling short-term predictions to reflect the influence of the wind field on the motion trend of the envelope.

[0033] For different tower cranes, the controller uses the speed difference of their respective envelopes as the relative speed to establish a relative motion model. This relative motion model is then used as the input for short-time trajectory prediction, enabling the subsequently obtained predicted envelope to simultaneously reflect the combined effects of the tower crane's own rotation, luffing, hoisting actions, and wind disturbances on the envelope's motion trend. This provides a continuous and interpretable dynamic basis for calculating the minimum safe clearance and determining risk within the prediction time window.

[0034] Step S6: Based on the relative motion model, perform short-time trajectory prediction on the three-dimensional working envelope of different tower cranes to obtain the predicted envelope within the prediction time window, and perform hysteresis compensation on the predicted envelope when the preset conditions are met.

[0035] In one implementation, the controller uses the rotational speed and its rate of change as the criterion for triggering hysteresis compensation. Let the... Tower crane timestamp The corresponding rotational speed is The rate of change of rotational speed is And set the rotational speed threshold as The threshold for the rate of change is Then, hysteresis compensation is triggered when the following formula is satisfied:

[0036] in It is obtained from the difference in rotational speed at adjacent sampling times. Based on the above criteria, the triggering condition for hysteresis compensation is deterministic, which facilitates the suppression of hysteresis errors in the predicted trajectory under high rotational conditions.

[0037] Short-term trajectory prediction includes the relative motion model established in step S5, the discrete time position recursive prediction of the representative points of the three-dimensional operation envelope within the prediction time window, and the generation of the prediction envelope based on the predicted position sequence of the representative points. The prediction envelope consists of combining the predicted position sequence of representative points with the geometric outline parameters of the three-dimensional operation envelope, and applying an expansion amount to the predicted position sequence of representative points to obtain a prediction envelope that covers the spatial occupancy range within the prediction time window. The expansion amount is used to characterize the spatial outline expansion caused by wind field disturbance and hook and hoist swing. The hysteresis compensation includes fitting the projection trajectory of the predicted envelope on a preset horizontal plane to obtain a fitted trajectory when the rotation speed meets the preset conditions, and then correcting the position of the predicted envelope based on the fitted trajectory to compensate for the prediction deviation caused by the hysteresis of the rotation drive response or the sampling delay.

[0038] Step S7: Calculate the minimum safe clearance between different tower cranes within the prediction time window, and compare the minimum safe clearance with the dynamic safe distance threshold to obtain the risk assessment result; output early warning instructions or control instructions based on the risk assessment result.

[0039] The minimum safety clearance is the minimum distance between the prediction envelopes of different tower cranes within the prediction time window. The minimum safety clearance is obtained by calculating the envelope distance for each discrete prediction time within the prediction time window and taking the minimum value. When calculating the minimum safety clearance, first apply uncertainty boundary expansion to the predicted envelope to obtain a conservative predicted envelope, and then calculate the minimum distance between the conservative predicted envelopes of different tower cranes as the minimum safety clearance. The risk assessment results include warning level and control level. When the minimum safety gap is less than the warning distance threshold, a warning command is output. When the minimum safety gap is less than the control distance threshold, a control command is output. Both the warning distance threshold and the control distance threshold are determined based on the dynamic safety distance threshold.

[0040] Example 2 This invention provides a wind field-adaptive method for avoiding collisions in multi-tower operations, the method comprising the following steps: I. Obtain the static parameters and installation position information of each tower crane in a multi-tower operation scenario. Static parameters include tower height, boom geometry, rear axle geometry, and foundation safety distance. Establish a unified spatial coordinate system based on the installation position information. Establishing the unified spatial coordinate system involves using a fixed reference point at the construction site as the origin and a horizontal plane as the reference plane to create a global coordinate system. The installation position information of each tower crane is represented as the coordinates of the tower center point within this global coordinate system. Static parameters include the maximum working radius of the boom and the maximum overhang of the rear axle. These static parameters are entered into the tower crane model parameter table and stored in the storage module during system deployment. More details... During the system deployment phase, a unified spatial coordinate system is established, and static parameters are entered and stored. The unified spatial coordinate system is denoted as OXYZ, where O is the origin, XY are the horizontal reference planes, and Z is the vertical direction. The origin O is selected as a fixed reference point for the construction site, and the coordinate axis directions and the horizontal reference plane are determined during deployment. The horizontal reference plane represents the horizontal reference surface of the tower crane's operating area. For the... For a tower crane, the position of its tower center point in a unified spatial coordinate system is represented as: The center point of the tower is the vertical projection center point of the tower base or the center of a standard section of the tower. The foundation safety distance is denoted as... This is used as a benchmark value for subsequent calculations of the dynamic safety distance threshold.

[0041] The installation location information is obtained through measurement. First, fixed reference points are calibrated at the construction site. Then, the planar position of the center point of each tower crane relative to the fixed reference points is measured, obtaining the coordinates of the tower center point in a unified spatial coordinate system. The equipment used for measurement includes a total station, satellite positioning measurement equipment, and laser rangefinder equipment. Alternatively, the installation location information can be pre-entered based on the overall construction layout data and verified and calibrated through measurement when the system goes online to reduce input errors.

[0042] Static parameters are acquired and stored during the system deployment phase. These parameters originate from the tower crane's manufacturer's technical data, equipment nameplate information, on-site acceptance data, or tower crane model parameter tables. They are entered into the controller via the human-machine interface and then stored in the storage module. Static parameters can also be read from pre-stored parameters and written to the storage module via the communication interface with the tower crane's electrical control system. Static parameters related to the tower crane's geometry are represented using a unified notation system, with the tower height parameter denoted as... The length of the crane boom is denoted as The length of the rear axle is recorded as Static parameters also include the maximum working radius of the boom and the maximum overhang of the rear axle, which are used to define the operating coverage area and the overhang range of the rear axle of the tower crane, thereby providing a definite geometric basis for subsequent three-dimensional coordinate calculation, three-dimensional operation envelope construction, and risk assessment.

[0043] To ensure coordinate consistency during tower crane operations, the origin, coordinate axis directions, and installation position information of each tower crane in a unified spatial coordinate system are stored after deployment and remain unchanged during subsequent operation. When the construction site reference point changes, or when a tower crane is relocated or reinstalled, the installation position information is remeasured and updated to ensure that the three-dimensional coordinates of the boom end, hook, and rear axle end calculated in subsequent steps are all within the same unified spatial coordinate system. This facilitates subsequent envelope construction, minimum safety clearance calculation, and risk assessment.

[0044] 2. During operation, real-time operating status parameters and wind field parameters for each tower crane are acquired at preset sampling periods. Real-time operating status parameters include slewing angle, amplitude, hook height, slewing speed, luffing speed, and hoisting speed; wind field parameters include wind speed and direction. All acquired real-time operating status parameters and wind field parameters are timestamped. Slewing speed is calculated from the difference in slewing angle between adjacent sampling periods, luffing speed from the difference in amplitude between adjacent sampling periods, and hoisting speed from the difference in hook height between adjacent sampling periods. Wind field parameters are collected by wind speed and direction sensors installed on the tower crane's structural components.

[0045] In one embodiment, during the operation, the controller uses a sampling period Real-time operating status data for each tower crane is collected, along with wind field parameters from the construction site. This data provides input for subsequent dynamic safety distance threshold generation, 3D operational envelope construction, and short-term trajectory prediction. For the... Tower crane, at timestamp The At each sampling time, the rotation angle, amplitude, and hook height are collected and recorded as follows: and At the same timestamp Wind speed and direction were collected at the corresponding sampling times and recorded as follows: and The aforementioned real-time operating status parameters and wind field parameters are all stored with corresponding timestamps, enabling subsequent calculations to be performed based on a unified time reference.

[0046] To obtain the velocity input required for the relative motion model, the slewing speed, luffing speed, and hoisting speed are calculated using the difference between adjacent sampling times. Tower crane timestamp The corresponding rotational speed is denoted as And calculate as follows:

[0047] No. Tower crane timestamp The corresponding amplitude speed is denoted as And calculate as follows:

[0048] No. Tower crane timestamp The corresponding lifting speed is denoted as And calculate as follows:

[0049] Using the above differential calculation method, the velocity parameter, angle parameter, and displacement parameter are within the same sampling period. Having a consistent time reference provides a continuous and computable dynamic basis for subsequently establishing relative motion models and making short-term trajectory predictions.

[0050] Wind field parameters are collected by wind speed and direction sensors installed on the tower crane structure. The wind speed and direction sensors output wind speed. With wind direction And send it to the controller. Wind direction By defining the reference direction using a unified spatial coordinate system, wind direction data can be compared with the slewing direction of the tower crane under the same angular reference. This allows for directional correction of the dynamic safety distance threshold and calculation of wind field impact in short-term trajectory prediction. Obtaining wind speed and direction from the tower crane's structural components more closely reflects the actual wind conditions of the tower crane, thereby improving the reliability and consistency of wind field adaptive collision avoidance judgment.

[0051] To ensure the comparability of data from multiple tower cranes at the same time, in one implementation, the controller can perform time consistency processing on the data collected by different tower cranes based on timestamps, so that the real-time operating status parameters entering the three-dimensional coordinate calculation and envelope construction correspond to the same target time point as the wind field parameters, thereby improving the accuracy of subsequent minimum safety clearance calculation and risk assessment.

[0052] 3. Based on static parameters, real-time operating status parameters, and a unified spatial coordinate system, calculate the three-dimensional coordinates of the boom end, hook, and rear axle end of each tower crane, and construct the three-dimensional working envelope of each tower crane based on the three-dimensional coordinates. Specifically, before calculating the three-dimensional coordinates of the boom end, hook, and rear axle end, the collected data of different tower cranes are time-aligned according to the timestamps carried in step two. When calculating the three-dimensional coordinates of the boom end, hook, and rear axle end of each tower crane, the center point of the tower body is used as the reference point for coordinate calculation. The coordinates of the center point of the tower body in a unified spatial coordinate system are superimposed with the horizontal displacement determined by the rotation angle and amplitude and the vertical displacement determined by the hook height to obtain the corresponding three-dimensional coordinates. When constructing the three-dimensional working envelope, the boom sweep envelope and the rear axle sweep envelope are determined according to the boom geometric parameters and the rear axle geometric parameters. The hook spreader envelope is determined according to the three-dimensional coordinates of the hook and the preset expansion amount. Then, the boom sweep envelope, the rear axle sweep envelope, and the hook spreader envelope are merged to construct the three-dimensional working envelope.

[0053] In one embodiment, after obtaining the installation location information, static parameters, and real-time operating status parameters of each tower crane, three-dimensional coordinate calculation and operation envelope construction are performed to uniformly represent the main spatial occupancy range of each tower crane during group tower operation as a spatial object under a unified spatial coordinate system. This facilitates subsequent dynamic safety distance threshold correction, short-term trajectory prediction, minimum safety clearance calculation, and risk assessment. To ensure the comparability of data from different tower cranes at the same time, the controller first performs time alignment processing on the collected data from different tower cranes based on the timestamps carried in step two, so that the slewing angle, amplitude, hook height, and corresponding speed parameters entering the three-dimensional coordinate calculation correspond to the same target time point as the wind field parameters.

[0054] For the first For tower cranes, the center point of the tower is used as the reference point for coordinate calculations. The position of this reference point in a unified spatial coordinate system is combined with the displacement determined by the slewing angle, amplitude, and hook height to determine the spatial positions of the boom end, hook, and rear axle end. The horizontal position of the boom end is determined by the boom length and slewing angle, while the vertical position is determined by the tower height parameter. Their three-dimensional coordinates in the unified spatial coordinate system are expressed as follows:

[0055] The horizontal position of the hook is determined by both the amplitude and the rotation angle, while the vertical position is determined by the hook height. Their three-dimensional coordinates in a unified spatial coordinate system are as follows:

[0056] The horizontal position of the rear axle end is determined by the rear axle length and the slewing angle, and its horizontal displacement direction is opposite to that of the boom end. Its vertical position is determined by the tower height parameter. Its three-dimensional coordinates in a unified spatial coordinate system are as follows:

[0057] After obtaining the three-dimensional coordinates of the boom end, hook, and rear axle end, a three-dimensional working envelope of the tower crane is further constructed. The three-dimensional working envelope characterizes the spatial occupancy of the tower crane in its current operating state and is used for subsequent minimum safety clearance calculations. The three-dimensional working envelope includes the boom sweep envelope, the rear axle sweep envelope, and the hook spreader envelope. The boom sweep envelope is determined based on the structural dimensions determined by the boom's geometric parameters and the sweep area formed by the rotational motion. The rear axle sweep envelope is determined based on the structural dimensions determined by the rear axle's geometric parameters and the sweep area formed by the rotational motion. The hook spreader envelope is determined based on the hook's three-dimensional coordinates, the spreader's external dimensions, and the outward expansion. The outward expansion characterizes the potential spatial expansion range of the hook spreader under wind disturbances or operational swaying. The boom sweep envelope, the rear axle sweep envelope, and the hook spreader envelope are merged to obtain the three-dimensional working envelope of the tower crane. In this way, each tower crane obtains a comparable operational envelope expression in a unified spatial coordinate system, providing a consistent data basis for subsequent dynamic safety distance threshold correction, short-time trajectory prediction, and minimum safety clearance calculation.

[0058] Fourth, the dynamic safety distance threshold is obtained by correcting the basic safety distance based on the wind field parameters, and the dynamic safety distance threshold is directionally corrected based on the relationship between wind direction and rotation direction.

[0059] After obtaining the wind field parameters, the controller dynamically corrects the foundation safety distance based on wind speed and direction, generating a dynamic safety distance threshold. It then makes directional corrections based on the relationship between wind direction and the tower crane's rotation direction, enabling the safety distance criterion to adaptively adjust with wind conditions. This improves the safety margin and early warning stability in multi-tower operation scenarios. The dynamic safety distance threshold is jointly determined by the foundation safety distance, wind speed correction, and directional correction. The wind speed correction reflects the increased swing range of the hook and lifting equipment and the enhanced expansion trend of the work envelope caused by increased wind speed. The directional correction reflects the risk differences resulting from the coupling of wind direction and rotation direction.

[0060] To suppress frequent threshold jumps caused by instantaneous wind speed fluctuations, the wind speed used for correction calculations is a weighted combination of instantaneous wind speed and the average wind speed over a preset time window. The average wind speed over the preset time window is calculated from multiple wind speed samples within that time window. The wind speed correction amount is determined using a segmented approach based on preset wind speed levels, with different correction magnitudes corresponding to different wind speed levels. This ensures that the dynamic safety distance threshold increases with increasing wind speed while maintaining stability.

[0061] To achieve directional correction, the controller first determines the rotation direction. The rotation direction is determined by the change in the rotation angle within adjacent sampling periods; when the rotation angle increases, it is determined as the first rotation direction, and when the rotation angle decreases, it is determined as the second rotation direction. Wind direction is defined using a reference direction in a unified spatial coordinate system, allowing comparison between wind direction and rotation direction at the same angular reference. The controller determines the directional correction amount based on the angle between the wind direction and rotation direction, and classifies it into three correction states: downwind correction, crosswind correction, and headwind correction, depending on the angle interval. Under downwind correction, the dynamic safety distance threshold is increased; under headwind correction, it is decreased; and under crosswind correction, the dynamic safety distance threshold remains unchanged. To further solidify the wind field's adaptive characteristics, both the downwind correction increment and the headwind correction decrement are related to the wind speed level. The downwind correction increment increases with increasing wind speed level, while the headwind correction decrement decreases or remains unchanged with increasing wind speed level. This significantly improves the safety margin for downwind operation under strong wind conditions, while preventing excessive reduction of the threshold and decreased safety constraints under headwind conditions.

[0062] In one implementation, the dynamic safety distance threshold is denoted as... The basic safety distance is denoted as In timestamps At any given time, the dynamic safety distance threshold is determined by the following relationship:

[0063] To obtain the combined wind speed used for wind speed correction, the combined wind speed is defined as follows: Instantaneous wind speed is The preset average wind speed for the time window is The weighting coefficient is and satisfy ,but

[0064] The average wind speed within the preset time window The wind speed was calculated from the sampled values ​​within the time window, and the number of sampling points within the time window was [number missing]. ,but

[0065] Wind speed correction The wind speed level is determined using a segmented approach. The wind speed level threshold is assumed to satisfy... The corresponding wind speed piecewise correction amount satisfies ,but

[0066] Directional correction amount The directional correction is determined based on the angle between the wind direction and the rotation direction, wherein the directional correction amount is based on the time of each tower crane. The directions of rotation are determined respectively. For the first... For tower cranes, the direction of rotation is determined by the difference in rotation angle between adjacent sampling times. When it is the first turning direction, when This is the second turning direction. The angle between the wind direction and the corresponding azimuth of the turning direction is defined as... Let the threshold for the downwind angle be... The crosswind angle threshold is And satisfy The directional correction amount is determined according to three correction states:

[0067] Among them, the tailwind correction increment Headwind correction and reduction Determined based on wind speed levels in segments. Assume the downwind level increment sequence satisfies... The headwind level reduction sequence satisfies ,but

[0068] Through the above methods, the dynamic safety distance threshold can be adjusted in segments under different wind speed levels, and directional coupling correction can be achieved under both tailwind and headwind conditions. Furthermore, the tailwind increment and headwind reduction change with the wind speed level, thereby significantly improving the tailwind safety margin under strong wind conditions and maintaining the rationality and stability of the threshold adjustment under headwind conditions. This provides a safety criterion basis for subsequent comparison of the minimum safety gap of the predicted envelope, which varies with the operating conditions.

[0069] 5. Filter the time series of the three-dimensional working envelope and establish a relative motion model based on slewing speed, luffing speed, hoisting speed, and wind field parameters. This relative motion model determines the instantaneous velocity components of the three-dimensional working envelope based on slewing speed, luffing speed, and hoisting speed, and introduces wind speed and direction corresponding to the wind field parameters as external disturbance terms into the instantaneous velocity components, thus obtaining a relative motion model for short-time trajectory prediction between different tower cranes. (Details follow...) After constructing the 3D operational envelope and obtaining wind field parameters, the controller filters the time series of the 3D operational envelope and establishes a relative motion model based on the filtered results to provide smooth, continuous, and computable input for subsequent short-term trajectory prediction. Because the hook and lifting equipment in multi-tower operations sway due to wind disturbances, and sensor sampling and calculation introduce jitter, directly using unprocessed data for prediction can easily lead to significant fluctuations in the prediction results within a short period, resulting in false alarms or frequent changes in control commands. Therefore, the controller first smooths the key boundary position sequence of the envelope using a time window, and then suppresses short-term abrupt changes through a limiting strategy to obtain a stable envelope sequence.

[0070] For the first Tower crane, with timestamp Taking the three-dimensional coordinates of the crane boom tip at time t as an example, the filtered three-dimensional coordinates are denoted as follows: , Using a length of The sliding time window averaging filter is calculated as follows:

[0071] The three-dimensional coordinates of the hook and the rear axle end are obtained using the same method to obtain the corresponding filtered coordinates. To suppress short-term anomalous changes, the amplitude of the filtered coordinate increment is limited so that the coordinate change between adjacent sampling times does not exceed a preset amplitude limit threshold, thereby obtaining a smooth coordinate sequence and a smooth envelope sequence for prediction.

[0072] After obtaining the smooth coordinate sequence, a directly calculable relative motion model is established. Taking the hook as the representative point of the envelope motion, the three-dimensional coordinates of the hook in a unified spatial coordinate system are determined by the amplitude, rotation angle, and hook height, and are expressed as:

[0073] The speed of the hook is determined by the rotational motion, luffing motion, and hoisting motion. Let the rotational speed be... The amplitude change speed is The lifting speed is Then, ignoring higher-order minor quantities, the velocity components of the hook in a unified spatial coordinate system can be expressed as:

[0074] To incorporate the influence of wind field into the relative motion model, wind speed and direction are decomposed into a horizontal wind speed component in a unified spatial coordinate system. Let the wind speed be... The wind direction is Where the wind direction is defined with reference direction in a unified spatial coordinate system, the horizontal wind speed component of the wind field in the unified spatial coordinate system is:

[0075] The effect of wind disturbance on the horizontal motion of the hook and lifting device can be equivalently represented by an additional horizontal velocity term. Let the wind disturbance coefficient be... Then, considering the wind field disturbance, the equivalent velocity component of the hook is:

[0076] The wind field disturbance coefficient is used to characterize the equivalent contribution of the wind field to the horizontal movement trend of the hook and lifting equipment. Its value can be set by empirical parameters or determined by on-site calibration.

[0077] For any two different tower cranes and The relative position of the hook is defined as follows:

[0078]

[0079]

[0080] Its relative velocity component is obtained from the equivalent velocity difference between the two, and is expressed as:

[0081] Through the aforementioned relative velocity expression, a relative motion model suitable for short-term trajectory prediction is obtained, enabling subsequent predictions to simultaneously reflect the combined effects of the tower crane's own rotation, luffing, hoisting actions, and wind disturbances on the motion trend. For the prediction of the three-dimensional working envelope, the controller can calculate the velocity components using the boom end, hook, and rear axle end as representative points in the aforementioned manner, and merge the predicted trajectories of these representative points within the prediction time window to obtain the motion trend input of the predicted envelope. This provides a directly calculable dynamic basis for the subsequent calculation of the minimum safe clearance and risk assessment of the predicted envelope.

[0082] 6. Based on the relative motion model, the short-time trajectory prediction of the three-dimensional working envelope of different tower cranes is obtained to obtain the predicted envelope within the prediction time window, and lag compensation is performed on the predicted envelope when the preset conditions are met.

[0083] The short-term trajectory prediction includes the relative motion model established in step five, which performs discrete time-based recursive prediction of the representative points of the three-dimensional operation envelope within the prediction time window, and generates the prediction envelope based on the predicted position sequence of the representative points. The prediction envelope consists of combining the predicted position sequence of representative points with the geometric outline parameters of the three-dimensional operation envelope, and applying an expansion amount to the predicted position sequence of representative points to obtain a prediction envelope that covers the spatial occupancy range within the prediction time window. The expansion amount is used to characterize the spatial outline expansion caused by wind field disturbance and hook and hoist swing. The hysteresis compensation includes fitting the projection trajectory of the predicted envelope on a preset horizontal plane to obtain a fitted trajectory when the rotation speed meets the preset conditions, and then correcting the position of the predicted envelope based on the fitted trajectory to compensate for the prediction deviation caused by the hysteresis of the rotation drive response or the sampling delay.

[0084] In one implementation, after establishing the relative motion model, the controller performs short-time trajectory prediction on the three-dimensional working envelope to obtain the predicted envelope within the prediction time window, which is used for subsequent minimum safety clearance calculation and risk assessment. The short-time trajectory prediction is implemented using a discrete-time recursive method, that is, the positions of representative points are updated incrementally at fixed steps within the prediction time window. The predicted position sequence of representative points is used to generate the predicted envelope. Representative points can be selected from the end of the boom, the hook, and the end of the rear axle, so that the prediction results can cover the main spatial occupancy changes in the tower crane's forward extension direction, the lifting area, and the rear axle extension area.

[0085] In one implementation, the prediction time window length is set as follows: The predicted step size is The prediction steps are: For the first A representative point of tower cranes, taking the hook as an example, is its timestamp. The three-dimensional position at the corresponding time is The equivalent velocity component is For the prediction step Corresponding prediction time The hook position is recursively predicted as follows:

[0086] in The predicted position sequence of the boom end and the rear axle end can be obtained using the same recursive method, or the corresponding equivalent speed can be obtained based on their respective geometric relationships and speed mappings before recursive prediction.

[0087] After obtaining the predicted location sequence of representative points, a prediction envelope is generated. The prediction envelope is used to cover the spatial range that the tower crane may occupy within the prediction time window. For each prediction step... The controller uses the predicted positions of each representative point as a geometric reference, maps the structural outline parameters of the three-dimensional working envelope to the corresponding predicted positions, and applies an expansion amount to form a conservative predicted envelope. The expansion amount is used to characterize the spatial outline expansion caused by wind field disturbances and hook and hoist sway, and it is related to the wind speed level.

[0088] In one implementation, the expansion amount is based on the combined wind speed. The wind speed levels are segmented. Assume the wind speed level thresholds satisfy... The segmented values ​​of the expansion quantity satisfy Then the expansion amount Determined by the following formula:

[0089] The above segmentation rules result in a greater outward expansion with higher wind speeds, making the predicted envelope more conservative under strong wind conditions. This improves the safety margin for determining the minimum safe gap and reduces the probability of missed alarms.

[0090] Taking the hook as a representative point as an example, let the outward expansion be... Then the hook envelope in the prediction step The radius of the outer expansion range is available as follows The spatial extension representation allows the predicted envelope to cover the possible sway range under wind field disturbances. The predicted outlines of the boom sweep envelope, the rear axle sweep envelope, and the hook and spreader envelope are merged to obtain the predicted step. The predicted envelope is calculated, and the sequence of predicted envelopes for each prediction step is output as the predicted envelope within the prediction time window.

[0091] To compensate for prediction bias caused by lag in the slewing drive response or sampling delay, lag compensation is applied to the prediction envelope when preset conditions are met. These preset conditions can be related to the slewing speed; lag compensation is triggered when the slewing speed exceeds a preset threshold or the rate of change of the slewing speed exceeds a preset threshold.

[0092] Let the first Tower crane timestamp The corresponding rotational speed is The rate of change of rotational speed is And set the rotational speed threshold as The threshold for the rate of change is Then, hysteresis compensation is triggered when the following formula is satisfied:

[0093] in It is obtained from the difference in rotational speed at adjacent sampling times. Based on the above criteria, the triggering condition for hysteresis compensation is deterministic, which facilitates the suppression of hysteresis errors in the predicted trajectory under high rotational conditions.

[0094] The hysteresis compensation is achieved by projecting the trajectory of the representative point of the predicted envelope onto a preset horizontal plane, fitting the projected trajectory to obtain the fitted trajectory, and then correcting the predicted position based on the fitted trajectory so that the predicted trajectory better conforms to the continuous change trend of rotational motion.

[0095] In one implementation, taking the horizontal projection of the hook-represented point as an example, let the horizontal projection coordinates be... For the prediction step A quadratic polynomial fit is performed on the set of projection points within the preset fitting window to obtain the fitting function:

[0096] in and The parameters are used for fitting. The controller corrects the projection position of subsequent prediction steps based on the fitting function, or reprojects the predicted projection position along the fitted trajectory to obtain the hysteresis-compensated projection trajectory. The hysteresis-compensated projection trajectory is combined with the corresponding vertical prediction position to obtain the hysteresis-compensated representative point prediction trajectory, and the prediction envelope is updated accordingly to improve the stability and accuracy of the prediction envelope under the turning condition.

[0097] 7. Calculate the minimum safe clearance between different tower cranes within the prediction time window, and compare the minimum safe clearance with the dynamic safe distance threshold to obtain the risk assessment result; output early warning instructions or control instructions based on the risk assessment result.

[0098] The minimum safety clearance is the minimum distance between the prediction envelopes of different tower cranes within the prediction time window. The minimum safety clearance is obtained by calculating the envelope distance for each discrete prediction time within the prediction time window and taking the minimum value. When calculating the minimum safety clearance, first apply uncertainty boundary expansion to the predicted envelope to obtain a conservative predicted envelope, and then calculate the minimum distance between the conservative predicted envelopes of different tower cranes as the minimum safety clearance. The risk assessment results include warning level and control level. When the minimum safety gap is less than the warning distance threshold, a warning command is output. When the minimum safety gap is less than the control distance threshold, a control command is output. Both the warning distance threshold and the control distance threshold are determined based on the dynamic safety distance threshold.

[0099] In one implementation, after obtaining the predicted envelope and dynamic safety distance threshold within the prediction time window, the controller calculates the minimum safe clearance between the predicted envelopes of different tower cranes and compares the minimum safe clearance with the dynamic safety distance threshold to obtain a risk assessment result and output a warning command or control command. The minimum safe clearance characterizes the minimum separation between the spatial occupancy ranges of two tower cranes at any time within the prediction time window, while the dynamic safety distance threshold characterizes the safety distance constraints that should be maintained under the current wind farm conditions and slewing conditions.

[0100] In one implementation, for the first Tower crane and the first Tower crane, at timestamp The prediction envelope of the discrete prediction steps within the prediction time window is obtained at the corresponding time, and the prediction step is denoted as . The corresponding predicted time is The first Tower crane in prediction step The predicted envelope is denoted as , will the Tower crane in prediction step The predicted envelope is denoted as Without considering the expansion of the uncertainty boundary, the prediction step The envelope spacing is defined as:

[0101] in and These are spatial points on the boundary of the corresponding predicted envelope. This represents the Euclidean distance. The minimum safety gap within the prediction time window is defined as the minimum of all prediction step intervals, expressed as:

[0102] Based on the above definition, the minimum safety gap can reflect the most dangerous approach situation that may occur at any time within the prediction time window.

[0103] To incorporate acquisition errors, time consistency errors, and uncertainties caused by wind field fluctuations into the safety assessment, the controller applies uncertainty boundary expansion to the prediction envelope in one implementation to obtain a conservative prediction envelope.

[0104] Let the uncertainty boundary expansion be... The uncertainty boundary expansion is determined by the acquisition error expansion. Time consistency error spread With wind field fluctuation expansion The result of superposition is:

[0105] Among them, the acquisition error spread The time consistency error spread is determined based on the accuracy specifications of the sensors used to obtain the slewing angle, amplitude, hook height, and wind speed and direction, or based on the maximum measurement residual obtained after on-site calibration. Determined based on the upper bound of the sampling period and communication delay; wind field fluctuation spread. The wind speed fluctuation range within a preset time window is determined. Let the maximum wind speed deviation within the preset time window be... And let the fluctuation mapping coefficient be... ,but

[0106] For the first Tower crane in prediction step Predicted envelope To impose on it The conservatively predicted envelope is obtained by spatially expanding the expansion amount, denoted as . For the first The same logic applies to tower cranes. Based on the conservatively predicted envelope, the envelope spacing is calculated, yielding:

[0107] Furthermore, the conservative minimum safety margin within the prediction time window is obtained:

[0108] The controller will The minimum safety gap in step seven is used for risk assessment, ensuring that the risk assessment still meets the conservative safety principle when uncertainty exists.

[0109] Let the dynamic safety distance threshold obtained in step four be... Then for the first Tower crane and the first The risk assessment of a tower crane can be expressed as follows:

[0110] in This indicates a risk of collision. This indicates no collision risk. The controller performs the above calculations on each tower crane combination in the multi-tower scenario and uses the risk assessment results to output warning or control commands.

[0111] To achieve tiered response, a warning distance threshold is set in one implementation method. With control distance threshold and satisfy The warning distance threshold and control distance threshold are determined based on the dynamic safety distance threshold, and are expressed as follows:

[0112] in It is a proportionality coefficient and satisfies .when When, the controller outputs a warning command; when At this time, the controller outputs control commands. These commands are used to limit or stop the slewing, luffing, or hoisting movements of the relevant tower crane to reduce the risk of collisions within the predicted time window.

[0113] Through the aforementioned minimum safety clearance calculation and comparison mechanism, risk assessment takes the predicted envelope as the object and combines it with uncertainty boundary expansion, so that the assessment result can simultaneously reflect the future movement trend of the tower crane, the wind field adaptive safety distance threshold, and the impact of uncertainty, thus forming a complete closed loop from wind field adaptive threshold generation to predicted envelope clearance comparison and then to graded output.

[0114] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A wind field-adaptive method for collision avoidance in multi-tower operations, characterized in that, The method includes the following steps: Step S1: Obtain the static parameters and installation position information of each tower crane in the multi-tower operation scenario. The static parameters include tower height parameters, boom geometry parameters, rear axle geometry parameters, and foundation safety distance. Establish a unified spatial coordinate system based on the installation position information. Step S2: During the operation, acquire the real-time operating status parameters and wind field parameters of each tower crane at a preset sampling period; the real-time operating status parameters include slewing angle, amplitude, hook height, slewing speed, luffing speed and hoisting speed, and the wind field parameters include wind speed and wind direction; Step S3: Based on the static parameters, the real-time operating status parameters, and the unified spatial coordinate system, calculate the three-dimensional coordinates of the boom end, hook, and rear axle end of each tower crane, and construct the three-dimensional working envelope of each tower crane based on the three-dimensional coordinates; Step S4: Correct the basic safety distance based on the wind field parameters to obtain a dynamic safety distance threshold, and make directional corrections to the dynamic safety distance threshold based on the relationship between wind direction and rotation direction; Step S5: Filter the time series of the three-dimensional working envelope, and establish a relative motion model based on the slewing speed, the amplitude change speed, the lifting speed and the wind field parameters; Step S6: Based on the relative motion model, perform short-time trajectory prediction on the three-dimensional working envelope of different tower cranes to obtain the predicted envelope within the prediction time window, and perform hysteresis compensation on the predicted envelope when the preset conditions are met. Step S7: Calculate the minimum safe clearance between different tower cranes in the predicted envelope within the predicted time window, and compare the minimum safe clearance with the dynamic safe distance threshold to obtain the risk assessment result; output the early warning command or control command according to the risk assessment result.

2. The method according to claim 1, characterized in that, In step S1, the unified spatial coordinate system is a global coordinate system established with the fixed reference point of the construction site as the coordinate origin and the horizontal plane as the reference plane, and the installation position information of each tower crane is represented as the coordinates of the center point of the tower body under the global coordinate system. The static parameters include the maximum working radius of the boom and the maximum overhang of the rear axle.

3. The method according to claim 1, characterized in that, In step S2, both the real-time operating status parameters and the wind field parameters obtained are timestamped. The rotational speed is calculated from the difference in rotational angle between adjacent sampling periods; The amplitude variation rate is calculated from the amplitude difference between adjacent sampling periods; The lifting speed is calculated from the difference in hook height between adjacent sampling periods; The wind field parameters are collected by wind speed and direction sensors installed on the tower crane structure.

4. The method according to claim 1, characterized in that, In step S3, before calculating the three-dimensional coordinates of the end of the boom, the hook, and the end of the rear axle, the collected data of different tower cranes are time-aligned according to the timestamp carried in step S2. When calculating the three-dimensional coordinates of the boom end, hook and rear axle end of each tower crane, the center point of the tower body is used as the reference point for coordinate calculation. The coordinates of the center point of the tower body in the unified spatial coordinate system are superimposed with the horizontal displacement determined by the rotation angle and the amplitude and the vertical displacement determined by the hook height to obtain the corresponding three-dimensional coordinates. When constructing the three-dimensional working envelope, the boom sweep envelope and the rear axle sweep envelope are determined according to the boom geometry parameters and the rear axle geometry parameters, and the hook spreader envelope is determined according to the three-dimensional coordinates of the hook and the preset outward expansion. Then, the boom sweep envelope, the rear axle sweep envelope and the hook spreader envelope are merged to construct the three-dimensional working envelope.

5. The method according to claim 1, characterized in that, In step S4, the dynamic safety distance threshold is obtained by adding the basic safety distance and the wind speed correction amount. The wind speed correction amount increases with the increase of wind speed and is segmented according to the preset wind speed level. The wind speed used to determine the wind speed correction amount is a weighted combination of instantaneous wind speed and the average wind speed of the preset time window.

6. The method according to claim 1, characterized in that, In step S4, the rotation direction is determined by the direction of change of the rotation angle within adjacent sampling periods. When the rotation angle increases, it is determined as the first rotation direction, and when the rotation angle decreases, it is determined as the second rotation direction. The directional correction includes determining the directional correction amount based on the angle between the wind direction and the rotation direction, and dividing the directional correction amount into three correction states: downwind correction, crosswind correction, and headwind correction, according to the range of the angle. Downwind correction corresponds to increasing the dynamic safety distance threshold, headwind correction corresponds to decreasing the dynamic safety distance threshold, and crosswind correction corresponds to keeping the dynamic safety distance threshold unchanged.

7. The method according to claim 1, characterized in that, In step S5, filtering the time series of the three-dimensional working envelope includes performing sliding time window filtering on the coordinate sequence representing the position of the end of the crane boom, the hook and the end of the rear axle, and performing amplitude limiting processing on the filtered coordinate sequence to suppress short-term abrupt changes caused by the swing of the hook and lifting device, thereby obtaining a smooth envelope sequence for establishing the relative motion model.

8. The method according to claim 1, characterized in that, In step S5, the relative motion model determines the instantaneous motion velocity components of the three-dimensional working envelope based on the slewing speed, luffing speed, and hoisting speed, and introduces the wind speed and wind direction corresponding to the wind field parameters as external disturbance terms into the instantaneous motion velocity components, thereby obtaining a relative motion model for short-term trajectory prediction between different tower cranes.

9. The method according to claim 1, characterized in that, In step S6, the short-time trajectory prediction includes the relative motion model established in step S5, the discrete time position recursive prediction of the representative points of the three-dimensional operation envelope within the prediction time window, and the generation of the prediction envelope based on the predicted position sequence of the representative points. The prediction envelope includes combining the predicted position sequence of the representative points with the geometric outline parameters of the three-dimensional operation envelope, and applying an expansion amount to the predicted position sequence of the representative points to obtain a prediction envelope that covers the spatial occupancy range within the prediction time window. The expansion amount is used to characterize the spatial outline expansion caused by wind field disturbance and hook and hoist swing. The hysteresis compensation includes fitting the projection trajectory of the predicted envelope on a preset horizontal plane to obtain a fitted trajectory when the rotation speed meets the preset conditions, and correcting the position of the predicted envelope based on the fitted trajectory to compensate for the prediction deviation caused by the hysteresis of the rotation drive response or the sampling delay.

10. The method according to claim 1, characterized in that, In step S7, the minimum safety gap is the minimum distance between the prediction envelopes of different tower cranes within the prediction time window. The minimum safety gap is obtained by calculating the envelope distance for each discrete prediction time within the prediction time window and taking the minimum value. When calculating the minimum safety gap, first apply uncertainty boundary expansion to the predicted envelope to obtain a conservative predicted envelope, and then calculate the minimum distance between the conservative predicted envelopes of different tower cranes as the minimum safety gap. The risk assessment result includes a warning level and a control level. When the minimum safety gap is less than the warning distance threshold, a warning command is output. When the minimum safety gap is less than the control distance threshold, a control command is output. The warning distance threshold and the control distance threshold are both determined based on the dynamic safety distance threshold.

Images