Scheduling method between anti-collision base stations of forklifts

Abstract

The invention discloses a method for realizing scheduling between forklift anti-collision base stations, and relates to the technical field of forklift anti-collision base station scheduling based on wireless positioning, and the method comprises the following steps: setting a Fi nal dynamic window as a fixed value MaxNum = N, and processing data through a three-level data filtering mechanism: grouping initial Fi nal data by using a Hash chain table; according to the method, the positioning data processing precision and reliability are improved through a three-level data filtering mechanism and an evidence theory fusion algorithm, scene dynamic self-adaption is achieved through a fuzzy logic threshold algorithm and a time-space joint index, and the conflict processing efficiency is improved and the system stability is guaranteed by means of a double-closed-loop control mechanism and a greedy + backtracking mixed algorithm. And meanwhile, the method has cross-scene adaptability and fault self-healing capability, not only solves the problem of positioning failure caused by i ndex conflict of multiple base stations, but also realizes high-precision positioning, low-cost deployment and intelligent operation and maintenance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of forklift anti-collision base station scheduling based on wireless positioning, in particular to a scheduling method between forklift anti-collision base stations. BACKGROUND

[0002] In the industrial warehouse and logistics scene, the positioning and anti-collision system of the factory forklift is a key technology to ensure work safety. The existing forklift positioning and anti-collision system mainly consists of positioning base stations, positioning tags (installed on forklifts or personnel) and management platform system software. It realizes real-time positioning of forklifts and personnel through the cooperative work of multiple base stations, and realizes the anti-collision warning function of people and vehicles based on positioning data.

[0003] However, the existing technology has significant defects when dealing with complex scenes of multiple base stations and multiple tags: when multiple base stations simultaneously determine information for multiple tags, data processing errors often occur due to repeated base station index. Specifically, when the base station receives the positioning data (such as the Final frame) returned by the tag, if different base stations are assigned the same index, it will cause the system to be unable to accurately distinguish the ranging data of each base station, and thus produce false positioning information. This data unreality directly weakens the effectiveness of forklift anti-collision measures, leading to frequent forklift collision accidents and failing to maximize work safety.

[0004] In view of this, a scheduling method between forklift anti-collision base stations is provided to overcome the above problems. SUMMARY

[0005] The purpose of the present application is to provide a scheduling method between forklift anti-collision base stations to solve the problems raised in the background art.

[0006] To solve the above technical problems, the present application provides a scheduling method between forklift anti-collision base stations, which includes the following steps:

[0007] Set the Final dynamic window to a constant value MaxNum=N, and process the data through a three-level data filtering mechanism:

[0008] Use a hash-linked list to group the initial Final data, quickly locate the data block of the base station with the same index;

[0009] Use a sliding time window algorithm to implement frequency statistics on the same index data within 100ms or 200ms;

[0010] Introduce a Kalman filter algorithm to smooth the index frequency of abnormal fluctuations.

[0011] Further, the index conflict detection threshold T is dynamically calculated by a fuzzy logic threshold algorithm, and the formula is:

[0012] T = 15 + 5 x log2(D + 1) + 3 x V x 0.1 + I x 2;

[0013] Wherein:

[0014] Base threshold item: 15: minimum safety detection frequency in the single base station scenario without interference, as the benchmark value for threshold calculation;

[0015] Base station density parameter: D: the number of base stations per unit area in the target area, i.e. D = total number of base stations / area;

[0016] Coefficient: 5: amplification of the impact of density on the threshold;

[0017] Tag moving speed parameter: V: real-time moving speed of the tag carrier;

[0018] Environmental interference level parameter: I: 5-level quantized environmental interference level.

[0019] Further, when the index repetition is detected, the physical position offset of the conflict base station is calculated by the spatial vector positioning method, and if the offset is > 3 meters, it is determined as an independent base station conflict, and the priority adjustment strategy is started, priority = base station signal strength x 0.6 + installation timestamp x 0.4.

[0020] Further, a greedy + backtracking hybrid algorithm is used to implement index reallocation, and the free value is preferentially allocated from the current maximum index + 1, and if the allocation fails, it is backtracked to the nearest non-conflict index interval.

[0021] Further, further, an industrial-grade general micro control chip (such as STM32H743VI) is integrated in the base station, and the 1MB cache built-in the chip is used to store the index conflict log in real time, and the 200μs level fast interrupt response is supported by the chip native interrupt controller, forming a hardware closed loop; the management platform uses a distributed real-time operating system, and embeds an LSTM network model to predict potential conflict base stations in advance through historical data, forming a software closed loop.

[0022] Further, by a time-space joint index algorithm, MaxNum = N is bound with index adjustment period soltTime = Nms, N includes but is not limited to 100ms, 200ms, and each soltTime is divided into soltNum = 20 index slots, forming a three-dimensional mapping relationship of data frame-soltTime-soltNum.

[0023] Further, the evidence theory fusion algorithm is adopted, base station signal strength, transmission delay, index repetition frequency are taken as evidence body, conflict confidence is calculated through D-S synthesis rule, index adjustment is triggered when the confidence is greater than 0.7, and the confidence can be set according to the actual situation.

[0024] Compared with the prior art, the beneficial effects of the application are:

[0025] 1. The data processing accuracy and reliability are improved:

[0026] The three-level data filtering mechanism (hash chain table grouping, sliding time window frequency statistics, Kalman filter smoothing processing) reduces the data misjudgment risk exponentially, avoids false signals caused by transient interference and multipath effect, and improves the tag positioning accuracy to meet the requirements of precision workshop operation.

[0027] The evidence theory fusion algorithm fuses signal strength, transmission delay and index repetition frequency, and only triggers adjustment when the conflict confidence is greater than 0.7, which significantly reduces the misjudgment rate compared with the existing single threshold value, and the confidence can be set according to the actual situation.

[0028] 2. Dynamic adaptive scene capability:

[0029] The fuzzy logic threshold algorithm dynamically calculates the conflict detection threshold T according to the base station density D, tag moving speed (V), and environmental interference level (I), for example, in a high-density warehouse scene (D=5, V=2m / s, I=3), T=22, and in an open low-interference scene (D=1, V=1m / s, I=1), T=12.8, which automatically adapts to different environments to avoid false triggering.

[0030] The time-space joint index algorithm binds MaxNum=N and soltTime=Nms, N includes but is not limited to 100ms, 200ms, to form a three-dimensional mapping relationship, which ensures the uniqueness of the index in high-speed moving scenes in medium-density workshops and other scenes.

[0031] 3. Conflict processing efficiency and system stability:

[0032] Double closed-loop control mechanism: hardware closed loop realizes 200us level interrupt response through special chip, software closed loop uses LSTM network to predict potential conflict base station in advance, so that the conflict processing time is greatly shortened.

[0033] The greedy + backtracking hybrid algorithm combines the priority adjustment strategy (priority = signal strength x 0.6 + installation timestamp x 0.4) to ensure that data transmission is not interrupted during index reallocation, and to avoid positioning failure caused by conflicts in the prior art.

[0034] 4. System scalability and intelligence:

[0035] Cross-scene adaptability can adapt to scenarios such as warehouses, workshops, open sites, etc. without manual configuration of parameters. The dynamic window mechanism reduces the energy consumption of low-conflict scenarios.

[0036] The fault self-healing capability automatically marks the fault base station and allocates a standby index, maintains the continuous operation of the system, and the conflict in the prior art may cause the positioning failure of the whole region. BRIEF DESCRIPTION OF DRAWINGS

[0037] Fig. 1 A base station index conflict detection and dynamic adjustment schematic diagram for a scheduling method between forklift anti-collision base stations of the present application;

[0038] Fig. 2 A multi-parameter collaborative scheduling mechanism architecture diagram for a scheduling method between forklift anti-collision base stations of the present application. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0040] Please refer to Figs. 1-2 The present application provides a technical solution:

[0041] Please refer to Figs. 1-2 An embodiment of a scheduling method between forklift anti-collision base stations is shown in the figure:

[0042] 1. Final dynamic window value and data preprocessing:

[0043] Based on the efficient UWB multi-base station ranging calling method, the Final dynamic window of the forklift anti-collision positioning technology is set to a fixed value (i.e. MaxNum=N), so that the number of each group of data returned by the base station is fixed to MaxNum. On this basis, a three-level data filtering mechanism is added:

[0044] Primary filtering: the initial returned Final data is grouped by using a hash linked list, the data block of the same index base station is quickly located, and the time complexity is optimized from O(n) to O(1);

[0045] Intermediate filtering: a sliding time window algorithm is used to count the frequency of the same index data within 100ms or 200ms, so as to avoid misjudgment caused by instantaneous interference;

[0046] Advanced filtering: Introduce Kalman filter algorithm to smooth the index frequency of abnormal fluctuations, eliminate false repeated signals caused by multipath effect.

[0047] It needs to be supplemented here:

[0048] In the warehouse high-density scenario, MaxNum=N can balance the data transmission efficiency and conflict detection accuracy, avoiding missing detection of index conflicts due to data volume fluctuations.

[0049] In view of the problems that the fixed threshold in the prior art cannot adapt to the dynamic environment, and the single filtering method is easy to be disturbed, the data structure (hash linked list), time series analysis (sliding window), and signal processing (Kalman filter) are innovatively combined across fields to form a multi-level data purification system. The prior art only relies on single-layer data verification, while the present method reduces the data misjudgment risk exponentially through the cascading effect of three-layer filtering, solving the problem of misjudgment caused by transient interference.

[0050] 2. Index conflict resolution mechanism based on adaptive threshold:

[0051] When searching the collected base station data, the index adjustment is realized through the following innovative steps:

[0052] 2.1. Threshold dynamic calibration: Abandoning the fixed threshold (such as the original setting of 20 times), the fuzzy logic threshold algorithm is adopted to dynamically calculate the threshold T according to the base station density (D), tag moving speed (V), and environmental interference level (I) three elements, the formula is:

[0053] T=15+5×log2(D+1)+3×V×0.1+I×2;

[0054] Among them:

[0055] Basic threshold item: 15: the minimum safe detection frequency in the single base station scenario without interference, as the benchmark value for threshold calculation;

[0056] Base station density parameter: D: the number of base stations deployed per unit area (100m 2 ) in the target area, i.e. D=total number of base stations / area (100m 2 ), value range: D≥1 (single base station scenario), typical scenarios:

[0057] Warehouse high-density scenario: D=5-10 (deploy 5-10 base stations per 100m 2 );

[0058] Workshop medium-density scenario: D=2-4;

[0059] Open-air site low-density scenario: D=1-2;

[0060] Using the logarithmic function log2(D+1) instead of linear superposition, the impact of base station density on index conflict probability presents a marginal effect of decreasing (the higher the density, the slower the conflict risk growth rate brought by new base stations);

[0061] Coefficient: 5: amplify the impact of density on threshold, for example, when D=5, log2(6)≈2.58, the contribution value is 12.9, which significantly raises the threshold in high-density scenarios, avoiding frequent false triggers.

[0062] Tag moving speed parameter: V: real-time moving speed of the forklift (tag carrier), calculated by UWB signal Doppler shift or obtained by vehicle-mounted sensors, value range: 0≤V≤5 (typical speed range of industrial forklifts);

[0063] Speed and conflict risk are positively correlated: the faster the forklift moves, the more base stations it passes through per unit time, and the higher the instantaneous probability of index repetition. The term 3×V×0.1 is equivalent to 0.3×V, i.e., each 1m / s speed contributes 0.3 of the threshold increment. For example, when V=5m / s, it contributes 1.5, and in low-speed scenarios (V=1m / s), it only contributes 0.3, reflecting the linear adjustment of speed on the threshold.

[0064] Environmental interference level parameter: I: comprehensive evaluation of environmental interference level through signal strength standard deviation (σ), multipath delay spread (τ), etc. It is quantified in 5 levels:

[0065] I=1: low interference (such as open space, σ<5dB, τ<10ns);

[0066] I=2: low-medium interference (ordinary workshop, 5dB≤σ<10dB, 10ns≤τ<20ns);

[0067] I=3: medium interference (metallic shelf area, 10dB≤σ<15dB, 20ns≤τ<30ns);

[0068] II=4: medium-high interference (dense equipment area, 15dB≤σ<20dB, 30ns≤τ<40ns);

[0069] I=5: high interference (strong reflection environment, σ≥20dB, τ≥40ns);

[0070] The linear term I×2 makes each interference level contribute 2 to the threshold increment, for example, high interference (I=5) contributes 10, and low interference (I=1) only contributes 2, directly reflecting the impact of interference strength on index misjudgment risk.

[0071] Space factor (base station density D), motion factor (speed V), and environment factor (interference I) are integrated through nonlinear combination (log + linear) to solve the problem that fixed threshold cannot adapt to dynamic scenarios.

[0072] Dimensional unification: After adjusting the coefficients, the parameters are dimensionally unified to "times" (consistent with the index repetition unit), ensuring that the physical meaning is interpretable.

[0073] Scenario adaptation example:

[0074] High-density warehouse scenario (D = 5, V = 2 m / s, I = 3):

[0075] T = 15 + 5 × log26 + 0.3 × 2 + 3 × 2 ≈ 22 times;

[0076] Open-air low-interference scenario (D = 1, V = 1 m / s, I = 1):

[0077] T = 15 + 5 × log22 + 0.3 × 1 + 1 × 2 = 15 + 5 + 0.3 + 2 = 22.3 times;

[0078] Automatic threshold adjustment to avoid false positives in low-density scenarios.

[0079] 2.2, Conflict positioning and classification: When index repetition is detected, the physical location offset of the conflict base station is calculated using the spatial vector positioning method. If the offset is greater than 3 meters, it is determined as an independent base station conflict, and the priority adjustment strategy is started (priority = base station signal strength × 0.6 + installation timestamp × 0.4).

[0080] 2.3, index reallocation strategy: Using a greedy + backtracking hybrid algorithm, it starts from the current maximum index + 1 to allocate free values, and if it fails, it backtracks to the nearest non-conflict index interval, ensuring that data transmission is not interrupted during the reallocation process.

[0081] It should be noted here that the example is:

[0082] Reference Fig. 1 When Anc2 and Anc4 are both index = 3 and the dynamic threshold reaches T = 22 (assuming D = 5 base stations, V = 2 m / s, I = 3), the system performs the following steps:

[0083] Spatial vector positioning is performed on the signal of Anc4, confirming that its physical distance from Anc2 is 4.5 meters, and it is determined as an independent conflict.

[0084] Calculate the priority of Anc4 (signal strength -65 dBm, installation timestamp 1678521000), priority

[0085] = (-65 * 0.01 * 0.6) + (1678521000 * 1e-9 * 0.4) = 6.58;

[0086] From index = 4, detect index = 4 free, adjust Anc4 to index = 4 and write to base station registry, while sending reallocation log to management platform.

[0087] When index = 3 appears 22 times (11 times for Anc4), the system quickly locates the 11 data of Anc4 through the hash linked list, and confirms 10 as valid conflict data through Kalman filtering, triggering index adjustment to the free position.

[0088] 3、System integration:

[0089] 3.1、Optimized integration of factory forklift positioning anti-collision system:

[0090] This scheduling method is applied to a positioning system composed of positioning base stations, positioning tags (terminals) and management platform system software, and adds a double closed-loop control mechanism based on existing technology:

[0091] Hardware closed loop: integrate an industrial-grade general micro control chip (such as STM32H743VI) in the positioning base station, store the index conflict log in real time through the 1MB cache built-in the chip, use the chip's original interrupt controller to support 200μs-level fast interrupt response, and form a conflict fast detection and response mechanism at the hardware level;

[0092] Software closed loop: the management platform uses a distributed real-time operating system (RT-Thread), and embeds a conflict prediction neural network model in the Final data processing module. Through historical index conflict data, the LSTM network is trained to predict potential conflict base stations in advance.

[0093] Dynamic window and index adjustment coordination mechanism: the existing technology only sets the Final window or adjusts the index, this method binds the Final window value (such as 16 data per group) and the index adjustment period (default 100ms) through the time-space joint index algorithm, forming a three-dimensional mapping relationship of "data frame-time slice-index slot", so that each base station has a unique identifier in the time-space dimension; through the time-space joint index algorithm, the Final window value (MaxNum = N) and the index adjustment period (soltTime = 100ms, 200ms) are bound, and within each soltTime, 20 index slots are divided, forming a three-dimensional mapping relationship of "data frame-soltTime-soltNum". Among them:

[0094] soltTime: defined as the time period of index adjustment (default 100 ms), used to control the refresh frequency of dynamic window;

[0095] soltNum: represents the number of unique index slots that can be allocated within each soltTime (default 20), the dynamic allocation of index is realized through the greedy + backtracking algorithm, both of which constitute the space-time index system together with MaxNum;

[0096] When the forklift moves in the medium-density workshop (D=3), the configuration of soltTime=100 ms and soltNum=20 can ensure that the index allocation of up to 20 base stations is processed every 100 ms, avoiding index conflict when moving at high speed (V=3 m / s);

[0097] Anti-interference strategy of multi-base station fusion: Unlike the single threshold judgment of existing technology, this method uses evidence theory fusion algorithm, taking base station signal strength, transmission delay, and index repetition frequency as evidence, and calculates conflict confidence through D-S synthesis rule. When the confidence is >0.7, index adjustment is triggered, the false positive rate is reduced, and the confidence can be set according to the actual situation.

[0098] It needs to be supplemented here that:

[0099] In the same area, there are five base stations of index0, index1, index2, index2, and index3. The optimization process of the forklift passing through is as follows:

[0100] A, the tag sends a poll signal, the base station returns a resp, and the tag sends the Final data to the base station;

[0101] B, after the management platform receives the Final data, it first predicts the repetition probability of index2 as 0.82 (threshold 0.7) through the LSTM model, and activates the conflict detection module in advance;

[0102] C, when receiving Final for the first time, the evidence theory fusion algorithm calculates the conflict confidence of index2 as 0.78, triggering the adaptive threshold algorithm (D=5, V=1.5 m / s, I=2, calculating T=15+5×log26+3×1.5×0.1+2×2≈23.

[0103] It also needs to be supplemented that:

[0104] Positioning accuracy: through dynamic window and adaptive index adjustment, the positioning accuracy of the tag is improved, meeting the operation requirements of precision workshops;

[0105] Conflict processing speed: After introducing LSTM prediction and evidence theory fusion, the index conflict processing time is shortened.

[0106] Cross-scene adaptability: Through the fuzzy logic threshold algorithm, the system can automatically adapt to different scenes such as warehouse (high-density base station), workshop (medium-density), and open site (low-density) without manual reconfiguration of parameters. Existing technologies require manual adjustment of thresholds.

[0107] Fault self-healing capability: When the base station frequently conflicts due to hardware failure, the system will automatically mark it as "maintenance" and temporarily assign a standby index to ensure the overall system does not interrupt. In existing technologies, a conflict base station may cause the entire region to fail positioning.

[0108] Energy optimization: Dynamic window mechanism makes the frequency of tag sending Final data linked with index conflict frequency, reducing energy consumption in low conflict scenarios. Existing technologies send data at a fixed frequency.

[0109] Software portability: Conflict detection algorithm is written in C language, compatible with mainstream industrial operating systems such as RT-Thread and FreeRTOS. The code size is about 5KB, easy to integrate into existing forklift anti-collision systems.

[0110] Cost control: The cost of single base station conversion is about 80 yuan (mainly for STM32 chip cost), which is 84% lower than the existing solution of dedicated conflict detection module (cost 500+ yuan), suitable for large-scale factory deployment.

[0111] Summary:

[0112] Based on the efficient UWB multi-base station ranging method, the Final dynamic window of forklift anti-collision positioning technology is set to a constant value, i.e. MaxNum=N. In the high-density warehouse scenario, it can balance data transmission efficiency and conflict detection accuracy, avoiding index conflict missed detection due to data volume fluctuations. Meanwhile, a three-level data filtering mechanism is added: primary filtering uses hash linked list to group the initial returned Final data, optimizing the time complexity from O(n) to O(1) to quickly locate the same index data block; intermediate filtering uses a sliding time window algorithm to count the frequency of the same index data within 100ms, avoiding false positives caused by transient interference; advanced filtering introduces Kalman filtering algorithm to smooth the abnormal fluctuation of index frequency, eliminating false repeated signals caused by multipath effect. Through cross-domain combination, a multi-level data purification system is formed, reducing the risk of data misjudgment exponentially.

[0113] In the aspect of index conflict resolution, an adaptive threshold-based mechanism is adopted. Instead of fixed threshold, the fuzzy logic threshold algorithm is used to dynamically calculate the threshold T according to the three elements of base station density D, tag moving speed V, and environmental interference level I. The formula is T = 15 + 5 × log2(D + 1) + 3 × V × 0.1 + I × 2. The base station density parameter adopts a logarithmic function to reflect the diminishing marginal effect, and the coefficient 5 amplifies the influence of high-density scenarios; the tag moving speed is positively correlated with the conflict risk, with a speed of 1 m / s contributing 0.3 to the threshold increment; the environmental interference level is quantified into five levels, with each level contributing 2 to the threshold increment, with the dimension unified as "times". For example, in a high-density warehouse scenario (D = 5, V = 2 m / s, I = 3), T = 22, and in an open low-interference scenario (D = 1, V = 1 m / s, I = 1), T = 12.8, automatically adapting to different scenarios. When index duplication is detected, the spatial vector positioning method is used to calculate the physical position offset, and if it is greater than 3 meters, it is determined as an independent conflict, and the priority is calculated as "base station signal strength × 0.6 + installation timestamp × 0.4", and the greedy + backtracking hybrid algorithm is used to redistribute the index, starting from the current maximum index + 1 to ensure uninterrupted data transmission. For example, Anc2 and Anc4 are both index = 3 and T = 22, after positioning and priority calculation, Anc4 is adjusted to index = 4.

[0114] During system integration, it is applied to a system composed of positioning base stations, positioning tags, and management platforms, and a double-loop control mechanism is added: the hardware loop integrates an industrial-grade general microcontroller chip (such as STM32H743VI), uses its built-in 1MB cache and native interrupt controller to achieve 200μs-level interrupt response; the software loop uses the RT-Thread system and embeds the LSTM network to predict potential conflict base stations in advance. Through the time-space joint indexing algorithm, MaxNum = N and index adjustment period soltTime = Nms are bound, N includes but is not limited to 100ms, 200ms, and each soltTime is divided into soltNum = 20 index slots, forming a three-dimensional mapping relationship, adapting to medium-density workshop scenarios. The evidence theory fusion algorithm is used to fuse signal strength, transmission delay, and index repetition frequency, and when the confidence level is greater than 0.7, the adjustment is triggered, reducing the false negative rate.

[0115] The method combines the time dimension final window value with the space dimension index to form a cross-dimension base station identification system, introduces machine learning prediction and multi-source evidence fusion, realizes the upgrade of conflict detection from "passive response" to "active prediction + intelligent decision", adopts low-cost general hardware, considers the feasibility of new technology and factory production, solves the problems of existing technology relying on high-cost hardware or lacking dynamic adaptability, improves the label positioning accuracy, shortens the conflict processing time, has cross-scene adaptability, fault self-healing ability, reduces energy consumption, enhances hardware compatibility and software portability, and is suitable for large-scale deployment.

Claims

1. A method for implementing scheduling between forklift anti-collision base stations, characterized in that: The following steps are involved: Set the Final dynamic window to a fixed value of MaxNum = N, and process the data through a three-level data filtering mechanism: Use the hash table to group the initial Final data and quickly locate the base station data blocks with the same index; Use the sliding time window algorithm to perform frequency statistics on the same index data within 100ms; The Kalman filter algorithm is introduced to smooth the abnormally fluctuating index frequency.

2. The method for implementing scheduling between forklift anti-collision base stations according to claim 1, characterized in that: The index conflict detection threshold T is dynamically calculated using the fuzzy logic threshold algorithm. The formula is: T=15+5×log2(D+1)+3×V×0.1+I×2; in: Basic threshold item: 15: Minimum safety detection frequency in a non-interference, single-base station scenario, used as the benchmark value for threshold calculation; Base station density parameter: D: the number of base stations deployed per unit area in the target area, that is, D = total number of base stations / area; Coefficient: 5: Amplify the effect of density on threshold; Tag moving speed parameters: V: real-time moving speed of the tag carrier; Environmental interference level parameters: I: 5-level quantitative environmental interference level.

3. The method for implementing scheduling between forklift collision avoidance base stations according to claim 2, characterized in that: When index duplication is detected, the physical position offset of the conflicting base station is calculated using the spatial vector positioning method. If the offset is greater than 3 meters, it is determined to be an independent base station conflict and the priority adjustment strategy is activated. Priority = base station signal strength × 0.6 + installation timestamp × 0.

4.

4. The method for implementing scheduling between forklift anti-collision base stations according to claim 3, characterized in that: A greedy + backtracking hybrid algorithm is used to implement index reallocation, giving priority to allocating idle values ​​starting from the current maximum index + 1. If the allocation fails, it backtracks to the nearest non-conflicting index interval.

5. The method for implementing scheduling between forklift collision avoidance base stations according to claim 1, characterized in that: The positioning base station integrates an industrial-grade general-purpose microcontroller chip, which uses the chip's built-in 1MB cache to store index conflict logs and utilizes the chip's native interrupt controller to support 200μs-level fast interrupt response, forming a hardware closed loop. The management platform uses a distributed real-time operating system and embeds an LSTM network model to predict potential conflicting base stations in advance through historical data, forming a software closed loop.

6. The method for implementing scheduling between forklift anti-collision base stations according to claim 5, characterized in that: Through the time-space joint indexing algorithm, MaxNum=N is bound to the index adjustment period soltTime=Nms, where N includes but is not limited to 100ms and 200ms. Each soltTime is divided into soltNum=20 index slots, forming a three-dimensional mapping relationship of data frame-soltTime-soltNum.

7. The method for implementing scheduling between forklift anti-collision base stations according to claim 1, characterized in that: The evidence theory fusion algorithm is adopted, and the base station signal strength, transmission delay, and index repetition frequency are used as evidence bodies. The conflict confidence is calculated through the DS synthesis rule. When the confidence is greater than 0.7, the index adjustment is triggered, and the confidence can be set according to the actual situation.

Images