Forklift safety online industrial monitoring management system
By integrating data, calculating margins, and employing tiered intervention strategies within the online industrial monitoring and management system for forklift safety, the system addresses the real-time monitoring and prediction of multi-dimensional status parameters in forklift safety monitoring systems. This enables a shift from passive alarms to proactive intervention, thereby enhancing the inherent safety and operational efficiency of forklift operations.
Patent Information
- Application Number
- CN202511538020.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-27
- Publication Date
- 2026-01-30
- Estimated Expiration
- 2045-10-27
AI Technical Summary
Existing forklift safety monitoring systems lack continuous, real-time monitoring and analysis of multi-dimensional status parameters, making it impossible to predict future risks. Furthermore, their intervention measures are limited and fail to achieve a balance between ensuring safety and maintaining operational efficiency.
The data fusion unit generates a real-time state vector, the margin calculation unit generates a comprehensive safety margin, the risk prediction unit predicts future safety margin values, and the intervention control unit executes a tiered intervention strategy to achieve a shift from passive alarm to proactive intervention.
It realizes the fusion and prediction of multi-source data in the online industrial monitoring and management system for forklift safety, and can take graded intervention measures before potential dangers occur, so as to ensure a dynamic balance between safety and operational efficiency.
Smart Images

Figure CN121426010A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of industrial vehicle safety control, in particular to a forklift safety online industrial monitoring and management system. BACKGROUND
[0002] In the current industrial vehicle application field, forklifts as the core material handling equipment, their operation safety is directly related to the safety of personnel, goods and equipment; in order to ensure safe operation, forklifts are usually equipped with basic alarm systems, but the design of these systems relies on the experience judgment and operation of the driver, or the alarm is triggered only after the dangerous state occurs, which belongs to passive response safety strategy; the traditional safety protection mode lacks continuous, real-time comprehensive monitoring and analysis of multi-dimensional state parameters in the forklift driving process, such as kinematic data, position and attitude, and load distribution, and it is difficult to identify the evolution trend of potential risks before danger occurs.
[0003] In the prior art, the safety monitoring means is usually single, and cannot effectively fuse heterogeneous data from different sensors to form accurate and quantitative evaluation of the dangerous state of the whole vehicle; the existing system generally lacks prediction ability, that is, it cannot infer the risk level in the future short time window according to the trend of the current state data, thus missing the best intervention opportunity; in addition, when the risk is identified, the traditional intervention measure is usually single and non-classified alarm prompt, and cannot take automatic closed-loop control intervention consistent with the emergency degree of the risk, so as to achieve dynamic balance between safety protection and operation efficiency; Therefore, how to provide a forklift safety online industrial monitoring and management system capable of fusing multi-source data and based on safety margin calculation and prediction, realizing passive alarm to classified active intervention, is a technical problem to be solved by those skilled in the art; The above information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, and therefore it can include information which does not constitute prior art known to those of ordinary skill in the art. SUMMARY
[0004] To solve the above technical problems, the present application discloses a forklift safety online industrial monitoring and management system, in particular, the technical scheme comprises: A forklift safety online industrial monitoring and management system comprises: A data fusion unit is used to collect kinematic data, position and attitude information and load distribution data of the vehicle, and to perform spatio-temporal alignment processing on the collected data to generate a real-time state vector; A margin calculation unit is used to calculate a plurality of independent safety margin components based on the real-time state vector, and to perform fusion processing based on the safety margin components to generate a comprehensive safety margin; a risk prediction unit configured to predict a future safety margin value based on the time series of the comprehensive safety margin, and compare the future safety margin value with a preset safety threshold and a preset warning threshold to generate a risk level; an intervention control unit configured to generate and execute a corresponding hierarchical intervention strategy in response to the risk level.
[0005] Preferably, the generation process of the real-time state vector is as follows: kinematic data is obtained through the vehicle data bus, position and attitude information is obtained by using a positioning system, and load distribution data is obtained by using a linear force sensor array; the obtained data is time-aligned by using a timestamp correction algorithm, and the aligned data is unified to the centroid coordinate system of the vehicle to generate the real-time state vector. Preferably, the real-time state vector includes vehicle lateral acceleration, yaw rate, minimum collision time, load resultant force vector, and load moment vector.
[0006] Preferably, the generation process of the comprehensive safety margin is as follows: a rollover stability margin is calculated; a load stability margin is calculated; an environmental collision avoidance margin is calculated; the rollover stability margin, the load stability margin, and the environmental collision avoidance margin are multiplied to generate the comprehensive safety margin.
[0007] Preferably, the rollover stability margin is used to quantify the proportional relationship between the recovery moment and the overturning moment of the vehicle, and is calculated based on the vehicle lateral acceleration obtained from the real-time state vector, and in combination with the pre-calibrated equivalent center of gravity height and wheelbase of the vehicle.
[0008] Preferably, the load stability margin is used to evaluate whether the load moment is close to the design load limit, and is calculated based on the load moment obtained from the real-time state vector, and in combination with the maximum allowable load moment which is preset as a safety threshold parameter.
[0009] Preferably, the environmental collision avoidance margin is used to represent the nonlinear growth characteristics of the collision risk, and is calculated based on the minimum collision time obtained from the real-time state vector, and in combination with the calibrated time sensitivity coefficient by using an exponential decay model.
[0010] Preferably, the generation process of the risk level is as follows: the time series of the comprehensive safety margin is first-order differentiated to estimate the margin change rate; based on the margin change rate, a future safety margin value is predicted by using a linear extrapolation method; when the future safety margin value is greater than the preset safety threshold, the risk level is determined to be in a safe zone; When the future safety margin value is less than or equal to the preset safety threshold and greater than the preset warning threshold, the risk level is determined as a warning zone; When the future safety margin value is less than or equal to the preset warning threshold, the risk level is determined as a dangerous zone.
[0011] Preferably, the hierarchical intervention strategy includes a first-level intervention strategy; the first-level intervention strategy is executed in response to the risk level being the warning zone; the first-level intervention strategy includes executing a power output limiting operation, executing a steering angle speed limiting operation, and executing a tactile feedback operation.
[0012] Preferably, the hierarchical intervention strategy includes a second-level intervention strategy; the second-level intervention strategy is executed in response to the risk level being the dangerous zone; the second-level intervention strategy includes executing an active gentle braking operation, executing a maximum steering angle limiting operation, and executing a motion locking operation of a mast or a fork.
[0013] Compared with the prior art, the present application has the following beneficial effects: 1. Compared with the prior art safety strategy relying on the experience of a driver or passive alarm, the forklift truck safety online industrial monitoring and management system disclosed by the present application constructs a complete technical link from multi-source data sensing to closed-loop control, and realizes the transition from passive safety to active safety. 2. The system comprises a data fusion unit, a margin calculation unit, a risk prediction unit, and an intervention control unit, and forms a complete closed-loop control loop; firstly, the data fusion unit can acquire kinematic data, position and attitude information, and load distribution data of a vehicle, and perform space-time alignment processing by using a time stamp correction algorithm, and finally unify the data to a vehicle mass center coordinate system to generate a high-fidelity and real-time state vector without delay; this process solves the technical problem of multi-source heterogeneous data fusion, and lays a solid data foundation for subsequent accurate risk assessment. 3. The margin calculation unit of the present application receives the real-time state vector and converts it into an intuitive and quantifiable safety index; the unit calculates three independent safety margin components, i.e., a rollover stability margin, a load stability margin, and an environmental collision avoidance margin, by parallel computing; the components are nonlinearly fused by using a product operation to generate a comprehensive safety margin; this multiplication fusion design conforms to the short board effect principle in a safety system, can more accurately and sensitively capture the critical dangerous state of the whole system, and solves the disadvantage of the traditional weighted summation method that cannot effectively reflect an extremely dangerous situation in one dimension. 4. The risk prediction unit of the present application continuously analyzes the time series of the comprehensive safety margin, estimates the margin change rate by first-order difference processing, and predicts the future safety margin value by using a linear extrapolation method; this predictive decision mechanism enables the system to foresee the safety margin decay trend before the danger actually occurs, thereby reserving valuable reaction time for the subsequent intervention control module. 5、The intervention control unit of the present application converts the predicted risk level into specific, hierarchical executable vehicle control actions; when the system is in the warning zone, a first-level intervention strategy is executed, through flexible measures such as limiting power output, steering angular velocity, and providing tactile feedback, aiming to guide the driver to avoid risks, and minimizing the impact on normal operation efficiency; when the system enters the danger zone, a second-level intervention strategy is forcibly executed, including active gentle braking, maximum steering angle limitation, and mast or fork action locking, to decisively and effectively prevent accidents; this hierarchical, closed-loop control logic realizes the dynamic balance between safety and operation efficiency, ensuring the intrinsic safety of the vehicle in complex industrial scenarios. BRIEF DESCRIPTION OF DRAWINGS
[0014] The present application will be further explained in connection with the accompanying drawings and embodiments: Figure 1 is a structural block diagram of the system of the present application. DETAILED DESCRIPTION
[0015] In order to make the purpose, technical scheme and advantages of the present application more clear and explicit, the present application will be further explained in detail below in connection with specific embodiments.
[0016] Example 1: Please refer to Figure 1 A forklift safety online industrial monitoring and management system, comprising: A data fusion unit for collecting kinematic data, position and attitude information, and load distribution data of the vehicle, and performing spatio-temporal alignment processing on the collected data to generate a real-time state vector; A margin calculation unit for calculating a plurality of independent safety margin components based on the real-time state vector, and performing fusion processing based on the safety margin components to generate a comprehensive safety margin; A risk prediction unit for predicting future safety margin values based on the time series of the comprehensive safety margin, and comparing and analyzing the future safety margin values with a preset safety threshold and a preset warning threshold to generate a risk level; An intervention control unit for generating and executing corresponding hierarchical intervention strategies in response to the risk level.
[0017] The present embodiment discloses a forklift safety online industrial monitoring and management system; the core purpose of the system is to continuously monitor and predict the running state of the forklift in multiple dimensions, to realize the transition from passive alarm to active intervention, so as to eliminate potential safety risks in the embryonic state without affecting normal operation efficiency; the system disclosed in the present embodiment logically constitutes a complete closed-loop control loop, characterized by comprising a data fusion unit, a margin calculation unit, a risk prediction unit, and an intervention control unit; The data fusion unit is used to provide unified, synchronized and high-precision state input for the whole system; the unit is responsible for collecting kinematic data, position and attitude information and load distribution data of the vehicle from multiple heterogeneous sensors; due to the different sampling frequencies and transmission delays of these data sources, the data fusion unit performs space-time alignment processing, and finally generates a real-time state vector which can comprehensively describe the instantaneous dangerous state of the forklift; The margin calculation unit is used to convert complex, multi-dimensional state information into a single safety index which is intuitive and quantifiable; the unit receives the real-time state vector generated by the data fusion unit, and based on the specific components in the vector, calculates a plurality of independent safety margin components for evaluating specific risk dimensions; then, the unit performs nonlinear fusion processing on these components to generate a comprehensive safety margin; this comprehensive index can sensitively reflect the risk of any single dimension; The risk prediction unit is used to realize the leading role of safety management, that is, to give a warning before the dangerous state actually occurs; the unit continuously receives and analyzes the time series of the comprehensive safety margin, and by predicting the trend of the future safety margin value in a short time window, obtains a future safety margin value; then, the unit compares and analyzes the predicted value with the preset safety threshold and warning threshold in real time, thereby dynamically dividing the risk level of the current vehicle; The intervention control unit is used to convert the results of risk assessment into specific and executable vehicle control actions, forming a closed loop; the unit automatically generates and executes a set of corresponding intervention strategies in response to the risk level output by the risk prediction unit; for example, flexible prompts are executed at a lower risk level, and mandatory vehicle dynamic parameter restrictions are executed at a higher risk level, thereby actively maintaining the vehicle within the safe operating envelope; Through the cooperative work of the above four units of data fusion, margin calculation, risk prediction and intervention control, the present application constructs a complete technical link from data perception to closed-loop control; compared with the passive safety strategy in the prior art which relies on the experience of the driver or gives an alarm after the danger occurs, the present system can predict the decay trend of the safety margin in advance, and take active intervention measures matching the risk level; this not only greatly improves the intrinsic safety of the forklift operation, but also maximizes the reduction of unnecessary impact on normal operation efficiency through graded intervention, achieving a dynamic balance between safety and efficiency.
[0018] Example 2: The generation process of the real-time state vector is as follows: kinematic data is obtained through the vehicle-mounted data bus, position and attitude information is obtained by using the positioning system, and load distribution data is obtained by using the line force sensor array; the timestamp correction algorithm is used to perform time alignment processing on the obtained data, and the aligned data is unified to the center of mass coordinate system of the vehicle to generate a real-time state vector; The real-time state vector includes vehicle lateral acceleration, yaw rate, minimum collision time, load resultant force vector, and load moment vector.
[0019] In the system of Embodiment 1, the generation process of the real-time state vector is specifically implemented; the purpose of this is to ensure that the generation process of the state vector is reliable, accurate and engineering-implementation, and to provide a high-quality data basis for all subsequent calculations; The specific workflow of the data fusion unit is as follows: Firstly, the kinematic data of the vehicle is obtained through the vehicle-mounted data bus of the forklift; at the same time, high-precision positioning system is used to obtain high-frequency position and attitude information of the forklift in the global coordinate system; in addition, the load distribution data is monitored and obtained in real time through the linear force sensor array installed on the forks or the mast. Before subsequent processing, the data fusion unit first checks the validity of the original data of each sensor and filters it to eliminate abnormal values that may be caused by sensor noise or communication interference; at the same time, the dynamic and static obstacle information around the vehicle is perceived by the laser radar or visual sensor on board; the data fusion unit processes the perception information, identifies potential collision objects, and predicts the motion trajectories of both parties within a certain period of time according to their current state and kinematic data of the vehicle, thereby calculating the minimum collision time . Secondly, in order to solve the problem of non-uniformity of multi-source data in time reference, this embodiment adopts a timestamp correction algorithm based on Kalman filtering to process the obtained data in time alignment, which can accurately estimate and compensate the transmission delay of each item of data, and ensure that all data correspond to the same time in logic; finally, in order to eliminate the influence of the change of the vehicle's own pose on data interpretation, all time-aligned data is unified to the center of mass coordinate system of the vehicle to generate the real-time state vector . In this embodiment, the real-time state vector is explicitly defined as a vector containing five core components: ; wherein, is the vehicle lateral acceleration; is the yaw rate; is the minimum collision time; is the load resultant force vector; is the load moment vector; wherein the yaw rate and the load resultant force vector are reserved for higher-level dynamic stability models or driving behavior analysis, and the subsequent calculations of this embodiment will focus on the core risk components.
[0020] By specifying the data collection source, assigning the processing algorithm and unifying the coordinate reference, the embodiment ensures that the generation of real-time state vector is high-fidelity and non-delayed; this method not only solves the technical problem of multi-source heterogeneous data fusion, but also provides direct and effective input for the subsequent accurate calculation of safety margin by defining the specific composition, greatly improving the accuracy and reliability of the overall system risk assessment.
[0021] Embodiment 3: The generation process of the comprehensive safety margin is as follows: The roll stability margin is calculated; The load stability margin is calculated; The environmental collision avoidance margin is calculated; And the roll stability margin, load stability margin and environmental collision avoidance margin are multiplied to generate the comprehensive safety margin.
[0022] The roll stability margin is used to quantify the proportional relationship between the recovery moment and the overturning moment of the vehicle, which is calculated based on the vehicle lateral acceleration obtained from the real-time state vector, and combined with the pre-calibrated equivalent center of gravity height of the vehicle and the wheelbase.
[0023] The load stability margin is used to evaluate whether the load generated torque is close to the design load limit, which is calculated based on the load torque obtained from the real-time state vector, and combined with the maximum allowable load torque preset as a safety threshold parameter.
[0024] The environmental collision avoidance margin is used to represent the nonlinear growth characteristics of collision risk, which is calculated based on the shortest collision time obtained from the real-time state vector, and combined with the time sensitivity coefficient calibrated, using an exponential decay model.
[0025] In the system of embodiment 1, the generation process of the comprehensive safety margin is implemented; this decomposition and fusion strategy has the inherent logic of decomposing a general safety concept into multiple risk dimensions with clear physical meaning and independent modeling, and fusing them through a nonlinear way that can reflect the one-vote veto safety principle, so as to more accurately assess the dangerous state of the whole system; The margin calculation unit receives the real-time state vector as input, and calculates three independent safety margin components in parallel, each of which is normalized to interval, where 0 represents extreme danger and 1 represents absolute safety; First, the roll stability margin is calculated. The roll stability margin , which is used to quantify the proportional relationship between the restoring moment and the overturning moment of the vehicle, is a key indicator for evaluating the vehicle's ability to resist lateral overturning; in order to ensure the real-time performance of the calculation, the embodiment adopts a quasi-static model which can effectively reflect the rollover risk under most working conditions for approximate calculation; the calculation is based on the lateral acceleration of the vehicle obtained from the real-time state vector , and combined with the pre-calibrated inherent physical parameters of the vehicle, i.e. the equivalent center of gravity height of the vehicle and the wheelbase of the vehicle ; the specific calculation formula is as follows: ; Wherein: is the absolute value of the lateral acceleration obtained from the real-time state vector ; and are the pre-calibrated inherent physical parameters of the vehicle through three-dimensional modeling or experimental methods; is the gravitational acceleration; the function in the formula is used for lower bound constraint, to ensure that when the vehicle is in a supercritical state, the margin is 0 and the negative value is not negative; The second step is to calculate the load stability margin ; The load stability margin is used to evaluate whether the moment generated by the current load is close to the design load limit of the fork or the gantry; the calculation is based on the load moment obtained from the real-time state vector , and combined with the maximum allowed load moment which is pre-set as a safety threshold parameter; the specific calculation formula is as follows: ; wherein: is the modulus of the real-time load moment obtained from the real-time state vector ; is a safety threshold parameter determined according to the design manual provided by the manufacturer or through finite element analysis; the design of the square term is to improve the sensitivity of the margin when it approaches the limit state; The third step is to calculate the environmental collision avoidance margin ; The environmental collision avoidance margin is used to represent the nonlinear growth characteristics of the collision risk as the time approaches; the calculation is based on the shortest collision time obtained from the real-time state vector , and combined with the time sensitivity coefficient calculated by the exponential decay model: ; wherein: To obtain from the real-time state vector Obtain from; This is a time sensitivity coefficient, derived from statistical analysis of a large amount of historical driving behavior data, and is a value that best reflects the driver's sense of urgency; this formula ensures a non-linear characteristic mapping of collision risk. The calculated rollover stability margin Load stability margin Collision avoidance margin with the environment Perform multiplication to generate a comprehensive safety margin. : ; By decomposing the overall safety issue into three orthogonal dimensions—rollover, load, and collision avoidance—and modeling them separately, this embodiment achieves a precise characterization of risk sources. More importantly, it generates data using a multiplicative fusion method rather than an additive fusion method. This causes the margin component in any dimension to approach zero, which will lead to a decrease in the overall safety margin. Approaching zero; this design conforms to the principle of the weakest link effect in safety systems, and compared with the traditional weighted summation method, it can more accurately and sensitively capture the critical danger state of the entire system.
[0026] Example 4: The risk level is generated as follows: the time series of the comprehensive safety margin is subjected to first-order differencing to estimate the margin change rate; based on the margin change rate, the future safety margin value is predicted using linear extrapolation. When the future safety margin value is greater than the preset safety threshold, the risk level is determined to be a safe zone; When the future safety margin value is less than or equal to the preset safety threshold and greater than the preset warning threshold, the risk level is determined to be the warning zone; When the future safety margin value is less than or equal to the preset warning threshold, the risk level is determined to be a danger zone.
[0027] In the system of Example 1, the process of generating risk levels was implemented in a specific way; the purpose of this is to establish an objective and dynamic risk decision-making mechanism that not only focuses on the current margin status, but more importantly, can predict its future evolution trend, thereby achieving the leading role of control intervention. The risk prediction unit receives the comprehensive safety margin output by the margin calculation unit. The time series data is analyzed; to achieve prediction, the time series is first-differenced to estimate the margin change rate; subsequently, based on the estimated margin change rate, linear extrapolation is used to predict the future safety margin value. This prediction model is a simplified first-order approximation of the system's dynamic behavior, with low computational cost and strong real-time performance. This linear model is suitable for relatively stable operating conditions. For more dynamic, sudden events, other nonlinear prediction models can be used as supplements. The specific prediction formula is as follows: ; in: and From The margin values obtained from the time series at the current and previous time points; The sampling time interval of the system; This is a preset prediction time window; Preset safety threshold and preset warning thresholds It was determined through statistical analysis of a large amount of driving simulation data under hazardous conditions, combined with relevant safety standards. The preset prediction time window... It was optimized after comprehensively considering the typical response time of the vehicle and the driver's reaction time, aiming to reserve sufficient reaction time for the intervention control unit; Based on the predicted future safety margin , and preset threshold and Compare and dynamically classify risk levels: When the future safety margin value Greater than the preset safety threshold At that time, the risk level was determined to be a safe zone; When the future safety margin value Less than or equal to the preset safety threshold And greater than the preset warning threshold At that time, the risk level was determined to be a warning zone; When the future safety margin value Less than or equal to the preset warning threshold At that time, the risk level was determined to be a dangerous area; By introducing a linear extrapolation prediction model based on the rate of change, this embodiment shifts the focus of risk assessment from the current state. Transferred to the future state This predictive decision-making mechanism enables the system to anticipate dangers, reserving valuable reaction time for subsequent intervention and control modules. This allows intervention measures to be executed smoothly and effectively before the safety margin is completely exhausted, thus achieving true proactive safety control.
[0028] Example 5: The hierarchical intervention strategy includes a first-level intervention strategy; the first-level intervention strategy is executed in response to the risk level being in the warning zone; the first-level intervention strategy includes executing a power output limiting operation, executing a steering angle speed limiting operation, and executing a haptic feedback operation.
[0029] The hierarchical intervention strategy includes a second-level intervention strategy; the second-level intervention strategy is executed in response to the risk level being in the danger zone; the second-level intervention strategy includes executing an active gentle braking operation, executing a maximum steering angle limiting operation, and executing a motion locking operation of the mast or the forks.
[0030] In the system of embodiment 1, the hierarchical intervention strategy is specifically implemented; this is to convert the abstract risk level into precise and effective vehicle control instructions, and to ensure that the intervention intensity matches the severity of the risk; When the intervention control unit receives a signal that the risk level is in the warning zone, it indicates that the system predicts a potential risk, and the first-level intervention strategy is executed at this time; the design goal of this strategy is to flexibly and suggestively guide the driver to avoid risks; the first-level intervention strategy specifically includes: executing a power output limiting operation, such as limiting the maximum acceleration to 70% of the normal value; executing a steering angle speed limiting operation, which guides the driver to make a more gentle turn by applying dynamic damping to the steering system; and executing a haptic feedback operation, which issues a somatosensory warning by activating a vibration motor in the seat or steering wheel; When the intervention control unit receives a signal that the risk level is in the danger zone, it indicates that the system predicts that a danger is about to occur or has occurred, and the second-level intervention strategy is executed at this time; the design goal of this strategy is to forcibly and decisively intervene in vehicle control to ensure safety; the second-level intervention strategy specifically includes: executing an active gentle braking operation, the system automatically applies a braking force, and the target deceleration applied is precisely limited to a range that will not cause secondary instability of the load; executing a maximum steering angle limiting operation, which dynamically calculates and sets a maximum safe steering angle allowed in the current working condition according to the current vehicle speed and load state; and executing a motion locking operation of the mast or the forks, which temporarily prohibits all mast or fork operations that may further worsen the stability of the vehicle; By designing two levels of intervention strategies that correspond precisely to the warning zone and the danger zone, this embodiment achieves on-demand allocation of intervention intensity; the flexibility of the first-level intervention ensures that the impact on operating efficiency is minimized when the risk is low; and the forcible intervention of the second-level intervention ensures that accidents can be decisively and effectively prevented at critical moments; this hierarchical and closed-loop control logic enables the entire system to work cooperatively with the driver, ultimately achieving the core technical goal of maintaining the vehicle within the safe operating envelope in complex industrial scenarios.
[0031] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application.
Claims
1. A forklift truck safety online industrial monitoring management system characterized by, The application relates to a vehicle safety risk prediction and intervention system. The application comprises: a data fusion unit for collecting kinematic data, position and attitude information and load distribution data of a vehicle, and performing space-time alignment processing on the collected data to generate a real-time state vector; a margin calculation unit for calculating a plurality of independent safety margin components based on the real-time state vector, and performing fusion processing on the safety margin components to generate a comprehensive safety margin; a risk prediction unit for predicting a future safety margin value based on a time sequence of the comprehensive safety margin, and comparing and analyzing the future safety margin value with a preset safety threshold and a preset warning threshold to generate a risk level; 2. The on-line industrial monitoring and management system for forklift safety according to claim 1, wherein, an intervention control unit for generating and executing a corresponding hierarchical intervention strategy in response to the risk level. The generation process of the real-time state vector is as follows: kinematic data is obtained through a vehicle-mounted data bus, position and attitude information is obtained by using a positioning system, and load distribution data is obtained by using a linear force sensor array; a timestamp correction algorithm is used to perform time alignment processing on the obtained data, and the aligned data is unified to a center of mass coordinate system of the vehicle to generate the real-time state vector.
3. The on-line industrial monitoring and management system for forklift safety according to claim 1, wherein, The real-time state vector comprises vehicle lateral acceleration, yaw rate, minimum collision time, load resultant force vector and load moment vector. The generation process of the comprehensive safety margin is as follows: a rollover stability margin is calculated; a load stability margin is calculated; an environmental collision avoidance margin is calculated; 4. The on-line industrial monitoring and management system for forklift safety according to claim 3, wherein, and the rollover stability margin, the load stability margin and the environmental collision avoidance margin are multiplied to generate the comprehensive safety margin.
5. The on-line industrial monitoring and management system for forklift safety according to claim 3, wherein, The rollover stability margin is used to quantify the proportional relationship between the recovery moment and the overturning moment of the vehicle, and is calculated based on the vehicle lateral acceleration obtained from the real-time state vector, and in combination with the pre-calibrated equivalent center of gravity height of the vehicle and the wheelbase of the vehicle.
6. The on-line industrial monitoring and management system for forklift safety according to claim 3, wherein, The load stability margin is used to evaluate whether the load moment is close to the design load limit, and is calculated based on the load moment obtained from the real-time state vector, and in combination with the maximum allowable load moment which is preset as a safety threshold parameter.
7. The on-line industrial monitoring and management system for forklift safety according to claim 1, wherein, The environmental collision avoidance margin is used to represent the nonlinear growth characteristics of the collision risk, and is calculated based on the minimum collision time obtained from the real-time state vector, and in combination with the calibrated time sensitivity coefficient by using an exponential decay model. The generation process of the risk level is as follows: the time sequence of the comprehensive safety margin is subjected to first-order difference processing to estimate the margin change rate; the future safety margin value is predicted by using a linear extrapolation method based on the margin change rate; when the future safety margin value is greater than the preset safety threshold, the risk level is determined to be in a safety zone; when the future safety margin value is less than or equal to the preset safety threshold and greater than the preset warning threshold, the risk level is determined to be in a warning zone; 8. The on-line industrial monitoring and management system for forklift safety according to claim 1, wherein, when the future safety margin value is less than or equal to the preset warning threshold, the risk level is determined to be in a danger zone. The hierarchical intervention strategy comprises a first-level intervention strategy; the first-level intervention strategy is executed in response to the risk level being in the warning zone, and the first-level intervention strategy comprises executing a power output limiting operation, executing a steering angle speed limiting operation and executing a tactile feedback operation.
9. The on-line industrial monitoring and management system for forklift safety according to claim 1, wherein, The hierarchical intervention strategy includes a secondary intervention strategy; the secondary intervention strategy is executed in response to the risk level being in the dangerous zone, and the secondary intervention strategy includes executing a proactive gentle braking operation, executing a maximum steering angle limitation operation, and executing a motion lock operation of a mast or a fork.
Citation Information
Patent Citations
Intelligent forklift management system and method based on AI collision avoidance
CN120039807A
Bridge construction hazard source identification and risk analysis method and system
CN120144983A
Forklift dynamic path planning method based on deep reinforcement learning
CN120489164A
Double-rocker-arm derrick monitoring system based on multi-sensor data fusion
CN120778165A
Quantum, biological, computer vision, and neural network systems for industrial internet of things
US20230176550A1
Cited By
A control method and system of a spraying robot based on Mecanum wheels
CN122210659A
A control method and system of a spraying robot based on Mecanum wheels
CN122210659B
