Automatic balancing forklift control method and system based on multi-source data

Abstract

The application relates to the field of automatic counterbalance forklifts, in particular to an automatic counterbalance forklift control method and system based on multi-source data. The method comprises the following steps: acquiring multi-source data of sensors; compensating left fork weighing values and right fork weighing values according to real-time posture angles of the forklift, and generating equivalent left weighing values and equivalent right weighing values in a horizontal reference system; calculating a static gravity center offset component and a dynamic inertia offset component; adding or weightedly summing the static gravity center offset component and the dynamic inertia offset component to obtain a dynamic gravity center offset; inputting the real-time gravity center offset into a balance controller, and judging whether the absolute value of the gravity center offset reaches or exceeds a preset stable safety threshold by the balance controller; if yes, triggering a balance control strategy. The application realizes high-precision and low-delay dynamic gravity center estimation by fusing hierarchical feature construction and physical constraints of multi-source data, so that the effect of dynamically adjusting the forklift control strategy is achieved.

Description

Technical Field

[0001] This application relates to the field of automatic counterbalanced forklifts, and in particular to an automatic counterbalanced forklift control method and system based on multi-source data. Background Technology

[0002] With the rapid development of intelligent manufacturing and intelligent logistics systems, automated counterbalanced forklifts (hereinafter referred to as "forklifts") are key material handling equipment in scenarios such as warehousing, ports, and manufacturing workshops. Their level of intelligence directly affects overall operational efficiency and safety. Traditional forklifts mostly rely on manual operation, which has problems such as high work intensity, slow response speed, and high safety risks. In recent years, with the development of sensor technology, edge computing, and control algorithms, automated forklifts are gradually evolving towards an integrated perception-decision-execution model.

[0003] However, in complex and dynamic operating environments, forklifts are highly susceptible to tilting of the vehicle's center of gravity when lifting heavy loads due to uneven load distribution, ground tilt, and sudden changes in acceleration, leading to rollovers or instability. This is especially true at high lifting heights (e.g., above 3 meters), where even minor disturbances in the center of gravity can cause significant overturning moments. Existing automated forklift control systems often rely on single-source data (such as tilt sensors or load cells) for stability assessment, lacking the ability to deeply integrate and dynamically analyze multi-source heterogeneous data. Furthermore, they typically assume a uniform load distribution at the fork center or rely solely on static calculations of the center of gravity using load cells and geometric models, neglecting dynamic factors such as load eccentricity, fork extension / retraction nonlinearity, and vehicle pitch coupling. In summary, existing technologies suffer from inaccurate dynamic estimation of the center of gravity, leading to the failure of balance control strategies. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides an automatic counterbalanced forklift control method and system based on multi-source data.

[0005] Firstly, this application provides an automatic counterbalanced forklift control method based on multi-source data, employing the following technical solution:

[0006] The automatic counterbalanced forklift control method based on multi-source data includes the following steps: acquiring multi-source data from sensors; the multi-source data includes fork weighing values, fork extension length, forklift attitude angle, lateral acceleration, and ground slope angle; the fork weighing values ​​include the weighing values ​​of the left fork and the right fork.

[0007] The forklift's real-time attitude angles are used to compensate for the weighing values ​​of the left and right forks, generating equivalent left and right weighing values ​​in a horizontal reference frame. The static center of gravity offset component is calculated, with its direction determined by the equivalent left and right weighing values. The eccentricity is calculated, and its amplitude is determined based on the eccentricity. The dynamic inertial offset component is calculated, with the dynamic disturbance coefficient determined by the ratio of lateral acceleration to gravitational acceleration. The effective center of gravity height is the sum of the fork extension length and the forklift chassis's inherent center of gravity height, and the product of the dynamic disturbance coefficient and the effective center of gravity height is the dynamic inertial offset component. The static and dynamic inertial offset components are added or weighted to obtain the dynamic center of gravity offset. The real-time center of gravity offset is input to the balance controller, which determines whether its absolute value reaches or exceeds a preset stable safety threshold. If it does, the balance control strategy is triggered.

[0008] Optionally, the method for calculating the eccentricity strength is as follows: the sum of the equivalent left weighing value and the equivalent right weighing value is taken as the total load, and the difference is taken as the load deviation; the eccentricity ratio is obtained based on the load deviation and the total load; the load activation factor is determined based on the ratio of the total load to the rated maximum load of the forklift; the square of the eccentricity ratio is multiplied by the load activation factor to obtain the eccentricity strength.

[0009] Optionally, in determining the load activation factor based on the ratio of the total load to the rated maximum load of the forklift, the load activation factor is equal to the ratio of the total load to the rated maximum load when the total load is lower than the rated maximum load, and otherwise the load activation factor is 1.

[0010] Optionally, the attitude angles include: a roll angle to reflect the left and right tilt of the vehicle body and a pitch angle to reflect the front and rear pitch of the vehicle body.

[0011] Optionally, the equivalent left fork weighing value is obtained by multiplying the pitch compensation factor by the left fork weighing value and then subtracting half of the roll correction; the equivalent right fork weighing value is obtained by multiplying the right fork weighing value by the pitch compensation factor and then adding half of the roll correction.

[0012] Optionally, the pitch compensation factor is the sum of the square of the pitch angle divided by two and one.

[0013] Optionally, the roll correction is equal to the product of the total load after pitch compensation and the roll angle; wherein pitch compensation refers to multiplying the sum of the weighing values ​​of the left fork and the right fork by the pitch compensation factor.

[0014] Optionally, determining the direction of the static center of gravity offset component based on the equivalent left weighing value and the equivalent right weighing value includes: when the equivalent right weighing value is greater than the equivalent left weighing value, the static center of gravity offset component is positive, indicating that the center of gravity shifts to the right; when the equivalent left weighing value is greater than the equivalent right weighing value, the static center of gravity offset component is negative, indicating that the center of gravity shifts to the left; otherwise, the static center of gravity offset component is zero.

[0015] Optionally, determining the amplitude of the static center of gravity offset component based on the eccentricity intensity includes: performing a square root operation on the eccentricity intensity to obtain a normalized eccentricity ratio; and multiplying the normalized eccentricity ratio by half the distance between the forks to obtain the amplitude of the static center of gravity offset component.

[0016] Secondly, this application provides an automatic counterbalanced forklift control system based on multi-source data, which adopts the following technical solution:

[0017] The automatic counterbalanced forklift control based on multi-source data includes a processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the aforementioned automatic counterbalanced forklift control method based on multi-source data.

[0018] The beneficial effects are: the above-mentioned automatic counterbalance forklift control based on multi-source data is generated into a computer program and stored in the memory so that it can be loaded and executed by the processor. Thus, the system can be made according to the memory and the processor, which is convenient to use.

[0019] This application has the following technical advantages:

[0020] 1. By integrating multi-source heterogeneous data such as fork weighing, attitude angles (roll and pitch angles), lateral acceleration, fork extension length, and ground slope angle, a three-layer progressive center of gravity offset model was constructed, which includes attitude compensation, eccentric decoupling, and dynamic disturbance fusion. This significantly improved the estimation accuracy of the lateral offset of the vehicle's center of gravity under complex conditions such as high lifting, driving on slopes, and sharp turns, and effectively reduced the risk of rollover caused by misjudgment of the center of gravity.

[0021] 2. It relies only on basic operations (addition, subtraction, multiplication, division, squaring, and square root) and fixed geometric parameters (such as fork width, chassis inherent center of gravity height, and rated maximum load), without the need to build complex physical models, perform online calibration, or introduce adjustable hyperparameters. The calculation process is a deeply serial data stream, which can run stably in cycles on an in-vehicle embedded system, meeting industrial-grade real-time requirements, while also possessing strong robustness and high interpretability.

[0022] 3. By introducing a combination design of load activation factor and nonlinear eccentricity ratio, the system can automatically suppress noise interference under light load and improve the sensitivity to eccentricity under heavy load or severe eccentricity, thereby realizing intelligent identification and response to stability risks under different load conditions.

[0023] 4. By setting a safety threshold based on the chassis half-width ratio, once the composite center of gravity offset exceeds the threshold, balance control strategies such as speed limit, lifting prohibition, and counterweight activation can be triggered immediately, thereby improving the operational safety and intelligence level of the automated forklift. Attached Figure Description

[0024] Figure 1 This is a flowchart of the automatic counterbalanced forklift control method based on multi-source data according to an embodiment of this application. Detailed Implementation

[0025] This application discloses an automatic counterbalanced forklift control method based on multi-source data, referring to... Figure 1 The process includes steps S1-S5, as detailed below:

[0026] S1: Acquire multi-source data from the sensor.

[0027] Multi-source data includes fork weighing values ​​and fork extension lengths. (Unit: m), forklift attitude angle, forklift lateral acceleration (Unit: m / s²) and ground slope angle (Estimated by IMU or LiDAR fusion). Fork load values ​​include the left fork load value. and the weighing value of the right fork (Unit: kg); Attitude angles include roll angle. and pitch angle (Unit: rad). The fork load reading reflects the distribution of load along the lateral direction of the forks. The fork extension length determines the load lever arm. Lateral acceleration characterizes dynamic inertial disturbances. The ground slope angle affects the direction of gravity, thus altering the effective reading of the load cell. Example: In a smart warehouse, an automated counterbalance forklift is moving a pallet of goods from area A to area B. Due to uneven stacking, the left side of the pallet is slightly heavier; the forklift needs to turn on a slope and raise the forks to a height of 3.5 meters. At this time, all five types of data are dynamically changing and none can be omitted.

[0028] In one embodiment, left and right load cells are installed below the fork support beam with a sampling rate of 100Hz; a wire-type displacement sensor is installed in the fork's embedded slide rail to measure the fork extension length with a sampling rate of 50Hz; and a six-axis IMU (Inertial Measurement Unit) is installed near the forklift frame's center of gravity to output the roll angle. Pitch angle The system uses triaxial acceleration with a sampling rate of 200Hz; it also uses a lidar scanner to scan the local ground in front of or below the forklift to obtain 3D point cloud data; based on this point cloud, it uses a plane fitting algorithm (such as least squares or RANSAC) to estimate the tilt angle of the local ground as the ground slope angle. The above point cloud processing and slope estimation methods are well-known technologies in this field and will not be described in detail here.

[0029] In one embodiment, all sensors are connected to the vehicle edge computing unit (ECU) via CAN bus or EtherCAT industrial Ethernet, employing a hardware timestamp synchronization mechanism to ensure that multi-source data is aligned on the timeline. The specific process is as follows: the ECU triggers a global synchronization signal every 10ms; upon receiving the synchronization signal, each sensor completes data acquisition and timestamps it within the next sampling cycle; the data is filtered (e.g., by moving average or low-pass filtering) and then packaged and uploaded to the main control program; the main program executes a center of gravity estimation algorithm every 20ms.

[0030] S2: Compensate the weighing values ​​of the left and right forks based on the real-time attitude angle of the forklift to generate equivalent left and right weighing values ​​in the horizontal reference system.

[0031] On a level surface, the load cell reading directly reflects the load distribution. However, on a slope or in a tilting position, the direction of gravity changes, leading to... It is no longer equal to the actual load component. For example, when the forklift tilts forward... At times, some load will "transfer" to the front wheels, causing the fork load to be lower than its rated value. If the original load is used directly... Calculating the centroid will introduce a systematic negative bias. Conventional algorithms assume... The load is constant and the attitude effect is ignored. This application uses the original weighing data. Attitude modulation decoupling is performed to restore the equivalent load distribution and equivalent left-side weight value in the horizontal reference frame. and equivalent right weighing value .

[0032] Specifically, the pitch compensation factor is calculated; the pitch compensation factor is the sum of the square of the pitch angle divided by two and one, a form derived from the vertical component of gravity. The second-order Taylor expansion approximation is applicable to small-angle conditions (such as roll angle). Pitch angle This is used to compensate for the attenuation of the weighing reading caused by pitch.

[0033] When the forklift tilts forward or backward, the component of gravity perpendicular to the fork plane decreases, resulting in a lower weighing value. Theoretically, the actual total load reading output by the load cells on the forklift forks should be lower. (The reading of the load cell changes with the pitch angle) As it increases, it decreases, and its change follows the pattern of a cosine function. Therefore, its reciprocal is approximately... It is used to "amplify" the readings to restore the true load.

[0034] The total load after pitch compensation Total load reading without attitude compensation ;

[0035] Calculate the roll correction; the roll correction is equal to the product of the total load after pitch compensation and the roll angle; where pitch compensation refers to multiplying the sum of the weighing values ​​of the left and right forks by the pitch compensation factor. The roll correction is calculated based on the roll angle and total load to offset the load distribution shift caused by tilt.

[0036] Specifically, the roll angle causes gravity to have components in the left and right directions, hence the theory of left and right load difference. ; This indicates the total mass of the entire vehicle (including load); Represents gravitational acceleration; using replace And using small angle approximation The roll correction amount is obtained. .

[0037] Multiply the left fork weighing value by the pitch compensation factor, and then subtract half of the roll correction to obtain the equivalent left weighing value; multiply the right fork weighing value by the pitch compensation factor, and then add half of the roll correction to obtain the equivalent right weighing value.

[0038] In one embodiment, the equivalent left-hand weight value and equivalent right weighing value This can be expressed mathematically as: ; .Will , Substituting into the above formula,

[0039] Right now ; .

[0040] S3: Calculate the static center of gravity offset component; where the direction of the static center of gravity offset component is determined based on the equivalent left and equivalent right weight values, the eccentricity intensity is calculated, and the amplitude of the static center of gravity offset component is determined based on the eccentricity intensity.

[0041] Obtaining the equivalent load in the horizontal reference frame and If it is used directly to calculate the center of gravity offset, the fork extension length will be ignored. The nonlinear amplification effect on the center of gravity shift. Even if the load eccentricity is constant, when When the mass is increased, the same eccentric mass will produce a larger overturning moment, which is equivalent to a larger lateral shift of the center of gravity.

[0042] In determining the direction of the static center of gravity offset component based on the equivalent left and equivalent right weighing values, the static center of gravity offset component is positive when the equivalent right weighing value is greater than the equivalent left weighing value, indicating that the center of gravity shifts to the right; the static center of gravity offset component is negative when the equivalent left weighing value is greater than the equivalent right weighing value, indicating that the center of gravity shifts to the left; otherwise, the static center of gravity offset component is zero.

[0043] The method for calculating eccentricity strength is as follows: The sum of the equivalent left and right weighing values ​​is taken as the total load, and the difference is taken as the load deviation; the eccentricity ratio is obtained based on the load deviation and the total load; the load activation factor is determined based on the ratio of the total load to the forklift's rated maximum load; the square of the eccentricity ratio is multiplied by the load activation factor to obtain the eccentricity strength. ; Indicates the load activation factor. It represents the nonlinear eccentricity ratio, and the eccentricity intensity reflects the degree of danger of the load distribution.

[0044] The load activation factor is determined based on the ratio of the total load to the forklift's rated maximum load. When the total load is lower than the rated maximum load, the load activation factor equals the ratio of the total load to the rated maximum load; otherwise, the load activation factor is 1. Specifically, the load activation factor... The load activation factor is used to adaptively adjust the eccentricity sensitivity according to the current load level, and its value range is [0,1]. Indicates the total load after attitude compensation ( This reflects the current actual handling quality. This indicates the rated maximum load of the forklift, a fixed geometric / performance parameter determined by the forklift model. This represents the minimum value function. It approaches 0 under low loads and is 1 under high loads, which naturally suppresses noise. Nonlinear eccentricity ratio ; This represents the non-linear eccentricity ratio, with a value range of [0,1]. For example, when the eccentricity ratio is 0.5 (medium eccentricity), When it is 0.8 (severe eccentricity), It shows accelerated growth.

[0045] Determining the amplitude of the static centroid offset component based on the eccentricity intensity includes: performing a square root operation on the eccentricity intensity to obtain the normalized eccentricity ratio. Normalize the eccentricity ratio Multiplying this by half the fork spacing yields the amplitude of the static center of gravity offset component. Amplitude ;

[0046] In one embodiment, the mathematical expression for the static centroid offset component can be: ; symbolic function Ensure the offset direction is correct (if the right side is heavier, the center of gravity will shift to the right). This indicates the fork width, which is the distance between the centers of the left and right load cells. The value is determined by the vehicle model. The left and right load cells are symmetrically distributed on both sides of the center line, and the fork spacing is... Therefore, the lever arm length at the overturning moment is When a forklift rolls (rotates around the axis of travel), gravity generates a torque that causes the forklift to tip over.

[0047] S4: Calculate the dynamic inertial offset component; wherein, the dynamic disturbance coefficient is determined based on the ratio of lateral acceleration to gravitational acceleration; the sum of the fork extension length and the inherent center of gravity height of the forklift chassis is taken as the effective center of gravity height, and the product of the dynamic disturbance coefficient and the effective center of gravity height is taken as the dynamic inertial offset component.

[0048] Dynamic lateral acceleration This will generate an equivalent inertial force. The effect on the height of the vehicle's center of gravity At this point, an additional overturning moment is generated. This moment can be equivalent to a virtual center of gravity shift. That is, satisfying: ;but ;in This indicates the total mass of the entire vehicle (including load); This represents gravitational acceleration. But in reality... The dynamic changes in fork lifting height and load distribution cannot be directly obtained, so this application uses... ( (Based on the inherent center of gravity height of the chassis) Reasonable approximation Fork extension length is strongly correlated with lifting height, while lifting height dominates. The change in dynamic inertial offset component. The negative sign indicates that the direction of acceleration is opposite to the direction of deflection.

[0049] S5: Add or weighted sum the static center of gravity offset component and the dynamic inertial offset component to obtain the dynamic center of gravity offset; input the real-time center of gravity offset to the balance controller, which determines whether its absolute value reaches or exceeds the preset stable safety threshold; if it reaches or exceeds the threshold, the balance control strategy is triggered.

[0050] In one embodiment, the mathematical expression for the dynamic center of gravity offset can be: The dynamic center of gravity offset is determined by attitude compensation and It consists of decoupled static eccentricity and dynamic disturbances that have been effectively height-corrected.

[0051] In other embodiments, the mathematical expression for the dynamic center of gravity offset can be:

[0052] ;in, Depend on Mapped to Sigmoid We obtained, among which, ; To adjust the parameter (can be set to an initial value of 0.5). This refers to the nominal height, which is typically taken as 50% to 70% of the forklift's rated maximum lifting height; or the high-frequency usage height derived from historical operational data analysis. For example, if the forklift's maximum lifting height is 4 meters and the chassis's inherent center of gravity height is 0.8 meters, then it can be set as follows: , or simplified (Referring only to the forks). Introduction back, It becomes a dimensionless normalization factor, which facilitates its integration with the acceleration term.

[0053] In one embodiment, the safety threshold can be the product of a safety factor and half the chassis width. The safety factor can be taken as 0.3 to 0.4 (i.e., allowing the center of gravity offset to not exceed half the chassis width). ).

[0054] In one embodiment, the control strategy may be: limiting the fork height, prohibiting further lifting, preventing an increase in overturning moment, or reducing the forklift travel speed; activating an active counterweight system (if equipped), such as moving the counterweight to the left or right to generate a reverse torque compensation.

[0055] This application also discloses an automatic counterbalanced forklift control system based on multi-source data, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the automatic counterbalanced forklift control method based on multi-source data according to this application.

[0056] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0057] In this application, the aforementioned memory can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as resistive random access memory (DRAM), dynamic random access memory (DRAM), static random access memory (SRAM), etc., or any other medium that can be used to store required information and can be accessed by an application program, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device.

[0058] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for automatic balancing of a forklift truck based on multi-source data, characterized in that, The method comprises the steps of: obtaining multi-source data of a sensor; the multi-source data comprises a fork weighing value, a fork extension length, a posture angle of the forklift, a lateral acceleration and a ground slope angle; the fork weighing value comprises a left fork weighing value and a right fork weighing value; compensating the left fork weighing value and the right fork weighing value according to the real-time posture angle of the forklift to generate equivalent left weighing value and equivalent right weighing value in a horizontal reference system; calculating a static gravity center offset component; wherein the direction of the static gravity center offset component is determined according to the equivalent left weighing value and the equivalent right weighing value, the eccentricity intensity is calculated and the amplitude of the static gravity center offset component is determined according to the eccentricity intensity; calculating a dynamic inertia offset component; wherein the dynamic disturbance coefficient is determined according to the ratio of the lateral acceleration and the gravitational acceleration; the sum of the fork extension length and the inherent gravity center height of the forklift chassis is taken as the effective gravity center height, and the product of the dynamic disturbance coefficient and the effective gravity center height is taken as the dynamic inertia offset component; adding or weighted summing the static gravity center offset component and the dynamic inertia offset component to obtain a dynamic gravity center offset; inputting the real-time gravity center offset to a balance controller to determine whether the absolute value of the real-time gravity center offset reaches or exceeds a preset stable safety threshold; if so, triggering a balance control strategy; the calculation method of the eccentricity intensity is that the sum of the equivalent left weighing value and the equivalent right weighing value is taken as the total load, and the difference is taken as the load deviation; the eccentricity ratio is obtained according to the load deviation and the total load; the load activation factor is determined according to the ratio of the total load and the rated maximum load of the forklift; the square of the eccentricity ratio is multiplied by the load activation factor to obtain the eccentricity intensity; in the determination of the load activation factor according to the ratio of the total load and the rated maximum load of the forklift, the load activation factor is equal to the ratio of the total load and the rated maximum load when the total load is lower than the rated maximum load, and the load activation factor is 1 otherwise; the posture angle comprises a roll angle for reflecting the left-right tilting state of the vehicle body and a pitch angle for reflecting the front-rear tilting state of the vehicle body; the compensation of the left fork weighing value and the right fork weighing value according to the real-time posture angle of the forklift to generate equivalent left weighing value and equivalent right weighing value in a horizontal reference system comprises: calculating a pitch compensation factor; calculating a roll correction amount; multiplying the left fork weighing value by the pitch compensation factor, and then subtracting half of the roll correction amount to obtain the equivalent left weighing value; multiplying the right fork weighing value by the pitch compensation factor, and then adding half of the roll correction amount to obtain the equivalent right weighing value.

2. The method of claim 1, wherein, the pitch compensation factor is the sum of the square of the pitch angle divided by two and one.

3. The method of claim 2, wherein, the roll correction amount is equal to the product of the total load after pitch compensation and the roll angle; wherein the pitch compensation refers to multiplying the sum of the left fork weighing value and the right fork weighing value by the pitch compensation factor.

4. The method of claim 3, wherein, in the determination of the direction of the static gravity center offset component according to the equivalent left weighing value and the equivalent right weighing value, when the equivalent right weighing value is greater than the equivalent left weighing value, the static gravity center offset component is positive, indicating that the gravity center is offset to the right; when the equivalent left weighing value is greater than the equivalent right weighing value, the static gravity center offset component is negative, indicating that the gravity center is offset to the left; otherwise, the static gravity center offset component is zero. ​ 5. The method of claim 1, wherein, The amplitude of the static gravity center offset component determined according to the eccentric strength comprises: performing square root operation on the eccentric strength to obtain a normalized eccentricity ratio; and multiplying the normalized eccentricity ratio by half of the distance between the forks to obtain the amplitude of the static gravity center offset component.

6. An automatically balanced counterbalanced forklift control system based on multi-source data, characterized by, Comprise: A processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, realize the automatic balancing forklift control method based on multi-source data according to any one of claims 1-5.

Images