Intelligent timing type automatic dust removal system for loader

Abstract

The present application relates to the technical field of loader operation data processing and automatic control, and discloses an intelligent timing type automatic dust removal system for a loader. The system accesses engine speed, driving speed, hydraulic action state and dust removal execution record through a processor, forms a working condition characteristic sequence from continuous operation data, identifies operation phases such as shovel loading, transportation, lifting, unloading, returning or idling, and accumulates a dust load virtual clock according to phase dust contribution. According to the operation phase sequence, an executable window is predicted, a dynamic timing queue of advance, immediate, delay or waiting is generated, and a dust removal instruction is output by a control output interface. The present application makes dust removal triggering be constrained by dust load and operation window at the same time, reduces advance or lag dust removal caused by fixed period, and keeps the state of ventilation, heat dissipation or air inlet channel controllable continuously.

Description

Technical Field

[0001] This invention belongs to the field of loader operation data processing and automatic control technology, specifically relating to an intelligent timed automatic dust removal system for loaders. Background Technology

[0002] When loaders are operating in conditions such as loading and unloading sand and gravel, earthmoving, and stockpiling ore, dust easily accumulates in the engine compartment, cooling channels, air intake channels, and filter surfaces. Existing dust control methods for loaders mostly use timers or vehicle controllers to output fixed-cycle commands. The controller starts accumulating running time after the vehicle is powered on, and after a preset time, it outputs a start command to the dust removal actuator. The dust removal actuator performs blowing, backflushing, or dust cleaning actions according to the preset duration. After the actions are completed, the timer resets and enters the next timing cycle. Some solutions use engine running status or vehicle start / stop status as simple enable conditions to avoid malfunctions when the machine is stopped, but their core control still relies on fixed-time interval triggering. Timers typically only record the duration of power-on, engine running time, or the cumulative duration after dust removal; the trigger time is determined by the preset cycle, and there is no data-level linkage between the dust removal action and the vehicle's operating cycle. This type of solution is not only simple to implement and has a short control link, but it can also complete periodic dust removal in scenarios where the dust environment is relatively stable and the operation cycle changes little. However, the timing logic itself does not distinguish between the operation phases of the loader, such as loading, transporting, lifting, unloading, returning, or idling, nor does it distinguish between the differences in the amount of dust entering and the deposition rate within the same running time.

[0003] To mitigate the incompatibility of fixed-cycle methods, existing technologies incorporate operating time, engine speed, fan operating status, or temperature changes into dust removal control. These solutions typically determine whether dust removal should commence after a timer reaches its trigger condition, based on a single operating parameter, or preemptively trigger dust removal when decreased heat dissipation capacity, filter clogging trends, or abnormal engine load are detected. Some control schemes preset multiple operating modes, shortening the dust removal interval in heavy-load mode and extending it in light-load mode. While these schemes add operational status judgment compared to a single fixed-time method, they primarily rely on instantaneous thresholds, mode switching, or fixed correction coefficients, with the judgment still limited to single-point data or preset mode tables. Increased engine speed may stem from loading resistance, hill starts, or short-term throttle changes; temperature changes may be influenced by ambient temperature, cooling airflow, and continuous load, making it difficult for a single parameter to consistently correspond to the degree of dust deposition. Existing technologies do not translate phase changes in continuous operating cycles into predictable dust removal execution windows, nor do they accumulate the contribution of each operating phase to dust deposition into a non-uniformly varying timing reference. Therefore, the controller still has difficulty determining whether the current dust load is close to the level requiring dust removal, and whether there is a low-conflict period suitable for performing dust removal actions in the near future.

[0004] The main technical problem with existing intelligent timed automatic dust removal control for loaders lies in the lack of a unified control logic for the timed triggering benchmark and the actual dust deposition load and operating phase. Fixed-time timing can only reflect the passage of natural time and cannot reflect the non-uniform increase of dust load under different operating phases; individual state parameter judgments can only reflect the local operating state at a certain moment and cannot determine whether the dust removal action will conflict with high-load phases such as loading and lifting. When the loader frequently switches between loading, transporting, unloading, and returning processes within a short cycle, the amount of dust deposition within the same preset cycle may vary significantly, and the suitable time window for dust removal also changes with the operating rhythm. When the dust load is high, the resistance of ventilation, heat dissipation, or air intake channels will gradually increase, and the fixed-time mode may continue to wait because the natural time has not yet arrived; when the dust load is low, the fixed-time mode may be triggered directly when the natural time arrives. If the controller still relies on a fixed cycle, it may trigger prematurely when the dust load is low, or wait for the timer to end when the dust load is close to affecting ventilation, heat dissipation, or air intake. It may also schedule dust removal actions during periods of intensive hydraulic action, high engine load, or large changes in vehicle speed, resulting in a mismatch between the timed dust removal control and the actual operating status of the loader. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent timed automatic dust removal system for loaders, which can solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An intelligent timed automatic dust removal system for a loader includes a processor, a memory, and a control output interface. The memory stores program instructions executed by the processor. The processor is configured to access existing operating data of the loader and form a working condition feature sequence by dividing engine speed, driving speed, hydraulic action status, and dust removal execution records into continuous time slices. A work phase sequence is generated based on the described working condition characteristic sequence; The dust load virtual clock is accumulated based on the dust deposition contribution corresponding to each operation phase; A dynamic timing queue is generated when the dust load virtual clock matches the predicted executable window, and dust removal control commands are output by the control output interface.

[0007] Preferably, when forming the working condition feature sequence, the processor calculates the engine speed change rate, driving speed range, hydraulic action holding state, hydraulic action switching sequence and cumulative running time after dust removal for each continuous time slice, and combines the load lifting relationship, speed direction relationship and hydraulic action sequence relationship between adjacent continuous time slices into a phase discrimination vector; The processor marks the continuous time slice as one of the phases of loading, transporting, lifting, unloading, returning, or idling based on the matching result of the phase discrimination vector and the historical operation cycle segment, and writes it into the operation phase sequence in chronological order.

[0008] Preferably, when the processor accumulates the dust load virtual clock, it reads the basic dust contribution factor bound to each phase category according to the phase category in the operation phase sequence, and multiplies or weights the basic dust contribution factor with the engine load level, driving speed level and hydraulic action intensity of the current time slice to obtain the dust contribution amount of the time slice. The processor adds the dust contribution of the time slice to the residual dust load retained in the previous dust removal cycle to generate a non-uniformly increasing virtual timing value, and uses the virtual timing value as the current value of the dust load virtual clock.

[0009] Preferably, before generating the dynamic timing queue, the processor predicts candidate executable windows in the next control cycle based on the end phase, phase duration, and adjacent phase transition relationship of the job phase sequence. The processor writes engine load level, hydraulic action conflict level, vehicle speed stability level and dust removal trigger urgency level to the candidate executable windows respectively, and sorts the current value of the dust load virtual clock with the level combination of each candidate executable window to form a timed queue item for immediate execution, early execution, delayed execution or waiting. The control output interface outputs the dust removal control command according to the timed queue item.

[0010] Preferably, when marking the consecutive time slices, the processor first merges several adjacent consecutive time slices into candidate job segments, and performs stability screening on the phase discrimination vectors within the candidate job segments; When a short-time jump vector exists in the candidate operation segment, the processor performs a consistent reconstruction based on the phase label of the previous candidate operation segment, the phase label of the next candidate operation segment, and the hydraulic action switching order, and writes the reconstructed phase label into the operation phase sequence. When the short-time jump vector is inconsistent with both the preceding and following phase labels, the processor sets the corresponding continuous time slice as a segment to be confirmed and prohibits the segment to be confirmed from participating in the candidate executable window prediction.

[0011] Preferably, after each dust removal control command is executed, the processor extracts a baseline segment before dust removal and a recovery segment after dust removal, and calculates at least one of the following: air volume change characteristics, channel resistance change characteristics, or engine load drop characteristics. The processor writes the difference between the post-dust removal recovery segment and the pre-dust removal reference segment into the dust removal recovery record, and corrects the basic dust contribution factor for the same working phase in the next dust removal cycle according to the dust removal recovery record. When the dust recovery record is lower than the dust recovery record level of the same historical operation phase, the residual dust load is retained to the next dust removal cycle according to the difference between the two.

[0012] Preferably, when forming the timing queue item, the processor establishes a window priority table jointly indexed by the operation phase, engine load level, hydraulic action conflict level, and vehicle speed stability level; The processor divides the current value of the dust load virtual clock into a safe interval, a near interval, and a trigger interval, and jointly encodes the sorting result of the window priority table with the safe interval, the near interval, or the trigger interval. When the joint encoding is a close interval and there is a high-priority candidate executable window, the dynamic timer queue writes an early execution item; When the joint encoding is a trigger interval and there is no qualified candidate executable window, the dynamic timer queue writes the deferred execution item and the longest waiting boundary.

[0013] Preferably, the processor configures the queue generation time, expected execution time, source continuous time slice, and current value of the corresponding dust load virtual clock for each early execution item, delayed execution item, or immediate execution item in the dynamic timing queue; When a new consecutive time slice enters the operating condition feature sequence, the processor recalculates the overlap relationship between the expected execution time and the updated candidate executable window, and marks queue items whose overlap relationship has changed as items to be rearranged. The items to be rearranged are rewritten into the dynamic time queue according to the joint encoding, while retaining the original continuous time slices.

[0014] Preferably, the processor divides the dust removal recovery record into recovery sample clusters according to the operation phase, the current value of the virtual clock of dust load before dust removal, and the dust removal continuity record; The processor extracts median recovery features and discrete recovery features within each recovery sample cluster, and performs difference calculation between the latest dust removal recovery record and its corresponding recovery sample cluster. When the difference calculation result continuously falls into the decay range, the processor reduces the dust removal and zeroing ratio of the corresponding operation phase and merges the unzeroed part into the residual dust load. When the difference calculation result returns to the stable range, the processor restores the historical update step size of the basic dust contribution factor.

[0015] Preferably, after the processor outputs the dust removal control command through the control output interface, it reads the dust removal execution feedback time, feedback continuity record, and execution completion flag, and establishes a write-back relationship with the expected execution time, source continuous time slice, and current value of the dust load virtual clock in the corresponding queue item; When the execution completion flag is inconsistent with the feedback continuous record, the processor pauses the timing reset of the corresponding queue item and writes the current value of the dust load virtual clock into the temporary hold state; When a qualified candidate executable window is confirmed in the next consecutive time slice, the temporary hold state re-participates in the joint encoding.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention forms a working condition feature sequence from the existing operating data of the loader into continuous time slices, and generates a working phase sequence from the working condition feature sequence. Then, based on the dust deposition contribution corresponding to the working phase, a virtual clock for dust load is accumulated, transforming the timing control reference from natural time to a calculated time constrained by both work intensity and dust load. Engine speed, driving speed, hydraulic action status, and dust removal execution records are processed in the same time slice, enabling the conversion of loading, transporting, lifting, unloading, returning, or idling operations into continuous phase data. The virtual clock for dust load increases non-uniformly with phase category, engine load level, driving speed level, and hydraulic action intensity. The controller then predicts candidate executable windows based on the end phase of the working phase sequence, phase duration, and adjacent phase transition relationships, and matches the current value of the virtual clock with the candidate executable windows to form a dynamic timing queue for immediate execution, early execution, delayed execution, or waiting. The generation of dust removal commands is simultaneously constrained by the dust accumulation state and the working phase window; dust removal actions are no longer determined solely by natural time, but by the dust load calculation results and a low-conflict window. When the virtual clock enters the approach interval and a high-priority candidate executable window appears, the controller can write the dust removal action into the early execution item; when the virtual clock enters the trigger interval but the current window is not qualified, the controller can write the delayed execution item and the longest waiting boundary. This process can reduce the premature or delayed dust removal caused by simple point-to-point triggering, making the dust removal action closer to the moment when the dust load reaches the control boundary, and reducing the situation where the dust removal action falls into the stage of dense hydraulic action, high engine load, or large vehicle speed change. This control method addresses the problem of mismatch between the timing trigger reference and the actual dust deposition load and operation phase, making it easier to limit the changes in dust resistance in ventilation, heat dissipation, or intake channels within a continuously controllable range.

[0017] 2. This invention also corrects the operational phase sequence by merging continuous time slices, filtering phase discrimination vector stability, and reconstructing short-time jump vector consistency, avoiding direct changes in phase labels due to instantaneous speed fluctuations, brief hydraulic action switching, or load disturbances. Pre-dust removal baseline segments and post-dust removal recovery segments are used to form dust removal recovery records. Airflow change characteristics, channel resistance change characteristics, or engine load decline characteristics participate in the subsequent correction of the basic dust contribution factor, allowing the dust contribution under the same operational phase to be adjusted according to the actual dust removal recovery degree. After extracting median and discrete recovery features from the recovery sample clusters, the latest dust removal recovery record can be compared with historical records of the same type. When the difference continuously falls within the decay range, the controller reduces the post-dust removal zeroing ratio and retains the residual dust load, preventing the unabridged dust from being mistakenly treated as zero load after the dust removal operation is completed. In the dynamic timing queue, a binding relationship is established between the queue generation time, the expected execution time, the source continuous time slice, and the current value of the dust load virtual clock. When a new continuous time slice enters, queue items with changing overlap relationships can be rearranged, ensuring that queue items update with the job phase while retaining the original source continuous time slice as a backtracking basis. The dust removal execution feedback time, continuous feedback recording, and execution completion marker are written back to the corresponding queue items. If feedback is inconsistent, timing is paused and reset, entering a temporary hold state, ensuring that dust removal control retains a traceable timing basis even in cases of data jumps, insufficient recovery, or inconsistent execution feedback. Through continuous constraints on residual dust load, recovered sample clusters, queue rearrangement, and feedback writing back, the controller can prevent residual dust load from being erroneously cleared, reduce the occurrence of queue items triggering as planned after job phase changes, and ensure that subsequent virtual clock accumulation continues to inherit the actual state of the previous dust removal cycle, maintaining a data closed loop between phase recognition, load accumulation, and command output. Attached Figure Description

[0018] Figure 1 The overall control flowchart of the intelligent timed automatic dust removal system for loaders; Figure 2 Flowchart for dust removal feedback correction and dynamic timed queue rearrangement. Detailed Implementation

[0019] refer to Figure 1In one embodiment, the intelligent timed automatic dust removal system for the loader includes a processor, a memory, and a control output interface. The memory stores program instructions executed by the processor. The processor reads engine speed, driving speed, hydraulic action status, and dust removal execution records through the loader's existing control link, without adding any new data collection objects unrelated to dust removal control. The processor buffers the continuously read data according to time slices of equal length or adaptive event boundaries, forming a working condition feature sequence for each time slice, including power status, driving status, hydraulic action status, and dust removal history status. The processor then identifies the current working phase of the loader based on the working condition feature sequence, whether it is loading, transporting, lifting, unloading, returning, or idling, and writes adjacent working phases into the working phase sequence in chronological order. Subsequently, it accumulates the dust load virtual clock according to the dust deposition contribution corresponding to each working phase. The dust load virtual clock does not change uniformly according to natural time, but instead forms a non-uniform increment value as the operation phase and operating load change. The processor predicts whether there is an executable window suitable for outputting dust removal control instructions in the next control cycle based on the changing trend at the end of the operation phase sequence, and generates a dynamic timing queue when the current value of the dust load virtual clock matches the executable window. The control output interface outputs dust removal control instructions to the existing dust removal execution components according to the dynamic timing queue. The execution record after the dust removal is completed is then written back to the memory and used as input data for the operating condition characteristic sequence of the next cycle. The advantage of this embodiment is that a closed-loop control chain of data reading, phase recognition, virtual timing, window prediction, queue output and record writing is formed by the processor, memory and control output interface, which transforms the fixed time triggering mode into a dynamic timing control mode constrained by both dust load and operation phase.

[0020] In this embodiment, the processor uses a unified data field representation to organize the data of consecutive time slices, so as to avoid the lack of time alignment basis when different running data enter the phase recognition process. Specifically, the processor uses the time slice number as an index to write the engine speed change, driving speed range, hydraulic action holding state, hydraulic action switching sequence, cumulative running time after the last dust removal, and dust removal execution record within the start and end times of the time slice into the same data row. When a short-term data loss occurs in a certain time slice, the processor retains the time index of that time slice and writes a pending confirmation mark to the missing field, instead of directly deleting the time slice, and then in the adjacent time slice... When the inter-slice has complete data, it is then supplemented or isolated. The operating condition feature sequence is composed of data rows from multiple consecutive time slices arranged in chronological order. The operation phase sequence is composed of phase tags corresponding to each consecutive time slice or merged candidate operation segments. The dust load virtual clock and the operation phase sequence share the same time index. Each queue item in the dynamic timing queue stores the source consecutive time slice, enabling subsequent dust removal feedback to trace back to the operating condition data corresponding to the trigger time. The advantage of this embodiment is that it forms a unified time slice index and field mapping relationship, avoiding data breakpoints caused by storing natural timing, operation phase, and dust removal feedback separately. For the field composition and processing meaning of the operating condition feature sequence, please refer to Table 1.

[0021] Table 1. Field Composition and Processing Meaning of Operating Condition Characteristic Sequences Time slice number Processor generation Record continuous running segments in chronological order. Operating condition characteristic sequence Engine speed change rate Existing running data Calculate the trend of rotational speed increase / decrease within the calculation time slice Phase discrimination vector driving speed range Existing running data Mark low speed driving, steady driving, or speed change. Phase discrimination vector Hydraulic action holding state Existing running data Mark lifting, lowering, holding, or no action. Phase discrimination vector Hydraulic action switching sequence Existing running data Record the sequence of adjacent hydraulic actions. Phase discrimination vector Cumulative runtime after dust removal Dust removal execution record Record the cumulative operation since the last dust removal. Dust load virtual clock Phase label Processor generation Mark loading, transporting, lifting, unloading, returning, or idling. Operation phase sequence Specifically, the processor represents the operating characteristics of the k-th consecutive time slice as a formula: ; in, Indicates the first The feature vector of the operating conditions for each consecutive time slice. Indicates the first The normalized value of the rate of change of engine speed within a continuous time slice. Indicates the first The encoded value of the driving speed range within a continuous time slice. Indicates the first The encoded value of the hydraulic action holding state within a continuous time slice. Indicates the first The encoded value of the hydraulic action switching sequence within a continuous time slice. Indicates the first The cumulative coded value of operation for a consecutive time slice relative to the end time of the previous dust removal cycle, for example, within a certain consecutive time slice. =0.60、 =0.30、 =0.80、 =0.70、 When the value is 0.50, the operating condition feature vector written by the processor is: This vector participates in subsequent phase recognition but is not directly used as a dust removal trigger condition.

[0022] In one embodiment, after forming the working condition feature sequence, the processor calculates the engine speed change rate, travel speed range, hydraulic action holding state, hydraulic action switching sequence, and cumulative running time after dust removal for each consecutive time slice. It also combines the load lifting / lowering relationship, speed direction relationship, and hydraulic action sequence relationship between adjacent consecutive time slices into a phase discrimination vector. Specifically, the processor divides the engine speed change rate into three calculation states: rising, stable, and falling; divides the travel speed range into three calculation states: low-speed movement, stable travel, and near-stop; encodes the hydraulic action holding state according to lifting holding, lowering holding, bucket action holding, and no hydraulic action holding; and follows the hydraulic action switching sequence of loading before lifting. The processor matches logical segments such as lifting before transporting, transporting before unloading, and unloading before returning. Then, it matches the field change relationships between adjacent time slices with historical operation cycle segments. If multiple consecutive time slices show low speed, increased load, and dense hydraulic action, it is marked as a loading phase. If multiple consecutive time slices show stable speed, hydraulic action maintained or low-frequency switching, it is marked as a transport phase. If the hydraulic action switches from transport maintenance to lifting or unloading, it is marked as a lifting or unloading phase. If the speed recovers and the hydraulic action decreases, it is marked as a return or idling phase. The advantage of this embodiment is that it identifies the operation phase by the causal sequence between adjacent time slices, reducing the situation where a single instantaneous data is directly interpreted as a dust removal trigger condition.

[0023] In this embodiment, the historical operation cycle segments are not fixed external templates, but are updated by the processor based on the marked operation phase segments during the continuous operation of the loader. Specifically, the processor stores multiple segment records in memory, each consisting of a phase tag, duration interval, engine speed change direction, driving speed change direction, and hydraulic action switching sequence. When a new candidate operation segment enters the identification process, the processor calculates the field consistency between the new candidate operation segment and each segment record, and selects the segment record with the highest consistency and satisfying the hydraulic action sequence as the phase tag source. If the consistency between the new candidate operation segment and all segment records is lower than acceptable, the processor marks the candidate operation segment as a segment to be confirmed and waits for subsequent adjacent segments to provide the phase relationship. Before confirmation, the segment to be confirmed does not participate in the executable window prediction, but only participates in the conservative accumulation of the dust load virtual clock. The conservative accumulation uses the higher dust deposition contribution in the confirmed adjacent phases as a temporary contribution source. The advantage of this embodiment is that the operation phase identification process has both historical segment support and pending confirmation protection, avoiding unconfirmed phases from directly changing the dust removal queue.

[0024] Furthermore, when accumulating the dust load virtual clock, the processor reads the basic dust contribution factor bound to each phase category according to the phase category in the operation phase sequence, and combines the basic dust contribution factor with the engine load level, driving speed level, and hydraulic action intensity of the current time slice for calculation. Specifically, the processor maintains basic dust contribution factors for the loading, transporting, lifting, unloading, returning, and idling phases respectively. The basic dust contribution factors are derived from the dust load growth records and post-dust removal recovery records of the same phase within historical dust removal cycles. The processor calculates the engine load level, driving speed level, and hydraulic action intensity for each time slice, wherein... The engine load level is represented by the relationship between the speed change rate and the load duration. The driving speed level is represented by the speed range and the degree of speed stability. The hydraulic action density is formed by the number of hydraulic action state changes and the switching direction within a unit time slice. The processor multiplies or weights the above levels with the basic dust contribution factor to obtain the dust contribution amount of the time slice. Then, it adds the contribution amount to the residual dust load retained in the previous dust removal cycle or connects it according to the zeroing ratio to generate a non-uniformly increasing virtual timing value. The advantage of this embodiment is that it transforms the difference in dust deposition in different working phases into a non-uniform change in the calculation time, avoiding the treatment of the same length of natural time as the same dust load.

[0025] The formula for calculating the virtual clock of dust load is as follows: ; in, Indicates the first The current value of the dust load virtual clock at the end of a continuous time slice. This represents the current value of the virtual clock representing the dust load at the end of the (k-1)th consecutive time slice. This represents the residual bearing capacity coefficient, which is set to 1 before dust removal and determined to be a value less than or equal to 1 after dust removal, based on the dust removal recovery record. This indicates the operation phase corresponding to the k-th consecutive time slice. Basic dust contribution factor This represents the normalized value of the engine load level. This represents the normalized value of the driving speed level. This represents the normalized value of hydraulic motion density. , and These represent the weights of engine load, driving speed, and hydraulic motion intensity in the dust contribution calculation, respectively. For example, at the end of the (k-1)th consecutive time slice... =12.00, when no dust removal occurred. =1, shovel phase =1.20, =0.75, =0.50, =0.80, =0.40, =0.20, =0.30, then This result serves as the input value for the next time slice to continue accumulating.

[0026] In one optional embodiment, before generating the dynamic timing queue, the processor predicts candidate executable windows for the next control cycle based on the end phase of the work phase sequence, the phase duration, and the transition relationship between adjacent phases. Specifically, the processor extracts multiple confirmed phases from the end of the work phase sequence, determines whether the current phase is at the end of the work cycle or a low-conflict segment, and combines the phase duration to determine whether the current segment is close to the phase switching boundary. If the end phase transitions from unloading to returning, or from returning to idling, or from the stable transport segment to the low-hydraulic action segment, the processor generates candidate executable windows. Subsequently, the processor writes a command to each candidate executable window. The system includes engine load level, hydraulic action conflict level, vehicle speed stability level, and dust removal trigger urgency level. The hydraulic action conflict level is determined by the hydraulic action holding state and switching number within the candidate window. The vehicle speed stability level is determined by the continuity of the speed range. The dust removal trigger urgency level is determined by the range in which the current value of the dust load virtual clock is located. The processor sorts multiple candidate executable windows to form a dynamic timing queue. The dynamic timing queue includes four types: immediate execution, early execution, delayed execution, and hold-wait. The advantage of this embodiment is that the window sorting and queue constraints are formed before the dust removal command is output, so that the dust removal action is no longer determined solely by whether the virtual clock reaches the trigger condition.

[0027] The priority value of the candidate executable window is determined according to the formula: ; in, Indicates the first The priority value of each candidate executable window. Indicates the first Normalized values ​​of vehicle speed stability level for each candidate executable window Indicates the first The phase adaptation level normalized value of each candidate executable window Indicates the first Normalized value of hydraulic action conflict level for each candidate executable window Indicates the first The normalized value of the engine load level for each candidate executable window. , , and These represent the weights of vehicle speed stability, phase adaptation, hydraulic action conflict, and engine load in the window sorting, for example, a candidate executable window. =0.80, =0.90, =0.20, =0.30, =0.30, =0.35, =0.20, =0.15, then The processor uses this priority value, along with the interval of the dust load virtual clock, for queue item type selection.

[0028] In this embodiment, the dynamic timing queue uses queue items as the basic control objects. Each queue item stores the queue generation time, the expected execution time, the source continuous time slice, the current value of the dust load virtual clock, the candidate executable window identifier, and the queue item type. The processor generates an early execution item when the dust load virtual clock enters the approach interval and a high-priority candidate executable window exists. It generates an immediate execution item when the dust load virtual clock enters the trigger interval and the current window meets the execution conditions. It generates a delayed execution item and writes it to the longest waiting boundary when the dust load virtual clock enters the trigger interval but the candidate executable window is not qualified. It generates a hold-wait item or does not generate an output instruction when the dust load virtual clock is in the safe interval. The control output interface outputs the dust removal control instruction only when there is an immediate execution item or an early execution item in the dynamic timing queue that has reached the expected execution time and has not been rearranged. If the delayed execution item reaches the longest waiting boundary and no qualified candidate executable window has appeared, the processor outputs the dust removal control instruction according to the conservative execution rule and records the source continuous time slice. The advantage of this embodiment is that it makes the early, immediate, delayed, and waiting operations all have a traceable data basis. For the queue item fields and processing rules of the dynamic timed queue, please refer to Table 2.

[0029] Table 2. Queue Item Fields and Processing Rules for Dynamic Timed Queues Queue generation time Time index for generating queue items Used to determine if a queue item has expired. Dynamic timed queue Expected execution time Predicting the time index of the executable window Used to control output interface triggering Control output interface Source: continuous time slice Data fragments generated by triggering queue items Used for feedback tracing Operating condition characteristic sequence Current value of virtual clock Dust load value when generating queue items Used for subsequent reset or retention Dust load virtual clock Queue item type Execute immediately, execute ahead of schedule, execute later, or wait. Used to determine the output order Dynamic timed queue Execution feedback flag Output, Completed, Pending Confirmation, or Error Retention Used for feedback writing Dust removal execution record Preferably, when marking consecutive time slices, the processor first merges several adjacent consecutive time slices into candidate operation segments, and performs stability screening on the phase discrimination vectors within the candidate operation segments. Specifically, the processor calculates the number of consecutive consistent values ​​for the engine speed change rate, driving speed range, and hydraulic action switching sequence within the candidate operation segments. When only a single field shows a short-term jump between adjacent time slices and the operation phase discrimination vectors of the preceding and following time slices still maintain the same trend, the processor marks the short-term jump vector as a reconstructable vector, and then, based on the phase label of the previous candidate operation segment and the phase of the next candidate operation segment... The label and hydraulic action switching sequence is reconstructed for consistency. If the short-term jump vector is inconsistent with both the previous and next candidate operation segments, the processor sets the corresponding continuous time slice as a segment to be confirmed. The segment to be confirmed is retained in the working condition feature sequence and continues to carry the time index, but does not participate in the prediction of the candidate executable window. The dust contribution corresponding to the segment to be confirmed is temporarily processed according to the higher contribution record among the adjacent confirmed phases until the subsequent phase relationship provides confirmation conditions. The advantage of this embodiment is that it isolates and reconstructs short-term data disturbances and avoids label breaks that are discontinuous with the actual operation cycle in the operation phase sequence.

[0030] In one optional embodiment, the processor uses a phase transition constraint table when performing consistent reconstruction of candidate job segments. The phase transition constraint table stores allowed adjacent phase relationships and prohibited adjacent phase relationships. Specifically, after the loading phase, it can transition to the lifting or transport phase; after the transport phase, it can transition to the lifting, unloading, or return phase; after the unloading phase, it can transition to the return or idling phase. If a short-term jump vector changes the phase label directly from loading to idling in a single time slice, and then reverts to loading or lifting in the next time slice, the processor considers the short-term jump vector as not satisfying the phase transition constraint and reconstructs it using the phase labels of the preceding and following segments. If no allowed transition relationship is satisfied in multiple consecutive time slices, the processor stops reconstruction and sets it as a segment to be confirmed. The segment to be confirmed is then rewritten into the job phase sequence based on the preceding and following phase relationships after a stable phase label appears. The advantage of this embodiment is that it introduces the sequential relationship of the loader's job cycle into the phase correction process, so that the phase recognition result is not dominated by a single jump data.

[0031] refer to Figure 2Furthermore, after each dust removal control command is executed, the processor extracts a pre-dust removal reference segment and a post-dust removal recovery segment. The pre-dust removal reference segment is the continuous running segment before the control output interface outputs the dust removal control command, and the post-dust removal recovery segment is the continuous running segment after the dust removal execution ends. The processor calculates at least one of the following characteristics from the two segments: airflow change characteristics, channel resistance change characteristics, or engine load reduction characteristics. When airflow or channel resistance feedback can be read from the loader's existing operating data, the processor uses airflow recovery and resistance reduction as recovery characteristics. When airflow or channel resistance feedback cannot be directly read, the processor uses similar characteristics. The engine load drop relationship under the working phase is used as a substitute recovery feature. The processor writes the difference between the recovery segment after dust removal and the reference segment before dust removal into the dust removal recovery record, and corrects the basic dust contribution factor of the same working phase in the next dust removal cycle according to the dust removal recovery record. When the dust removal recovery record is lower than the dust removal recovery record level of the same historical working phase, the processor does not completely reset the dust load virtual clock, but retains the residual dust load to the next dust removal cycle according to the difference ratio. The advantage of this embodiment is that the timing starting point after dust removal is determined by the actual recovery degree, avoiding the omission of residual load caused by directly clearing to zero after the execution of the action.

[0032] Dust removal recovery records are recorded according to the following formula:

[0033] Indicates the first Dust removal recovery record corresponding to each dust removal operation. Indicates the first Normalized value of airflow change in the recovery segment after dust removal relative to the baseline segment before dust removal. Indicates the first Normalized value of channel resistance change of the recovered segment after dust removal relative to the baseline segment before dust removal. Indicates the first Normalized value of engine load drop in the post-dust removal recovery segment relative to the pre-dust removal baseline segment. , and These represent the weights of changes in airflow, channel resistance, and engine load fallback in the recovery record, respectively. For example, the first... Secondary dust removal =0.70, =0.60, =0.50, =0.40, =0.35, =0.25, then The processor writes the result into the dust recovery record and uses it for subsequent correction of the residual dust load retention ratio and the basic dust contribution factor.

[0034] In this embodiment, the correction of the basic dust contribution factor does not use the method of directly overwriting the historical record with a single dust removal recovery record. Instead, the dust removal recovery record is bound to the corresponding operation phase, the current value of the virtual clock of the dust load before dust removal, and the dust removal continuity record. Specifically, when the dust removal recovery record is higher than the historical recovery record of the same phase, the processor reduces the residual bearing coefficient corresponding to the same operation phase in the next cycle and maintains the historical update step size of the basic dust contribution factor. When the dust removal recovery record is lower than the historical recovery record of the same phase, the processor increases the residual dust load retention ratio and slightly increases the basic dust contribution factor of the same operation phase, so that the virtual clock of the subsequent dust load approaches the approach range earlier under the same phase. If the dust removal recovery record is continuously lower than the historical recovery record of the same phase, the processor marks the corresponding dust removal execution record as an insufficient recovery record and increases the dust removal trigger urgency level during the dynamic timing queue generation process. The advantage of this embodiment is that it feeds back the dust removal execution result to the virtual timing model, so that the dust load of the next cycle accumulates and bears the actual recovery state of the previous cycle.

[0035] In one optional embodiment, when forming a timing queue item, the processor establishes a window priority table indexed by the job phase, engine load level, hydraulic action conflict level, and vehicle speed stability level. This window priority table is not used as a fixed trigger table but rather as an index structure for sorting candidate executable windows. Specifically, the processor uses the current job phase as the first index, the engine load level as the second index, the hydraulic action conflict level as the third index, and the vehicle speed stability level as the fourth index. These four indexes all point to a window priority value range. The processor then divides the current value of the dust load virtual clock into a safe range. The system considers the proximity interval and the trigger interval, and combines the sorting result of the window priority table with the interval where the dust load virtual clock is located. When the joint encoding corresponds to the proximity interval and there is a high-priority candidate executable window, the dynamic timing queue writes an early execution item. When the joint encoding corresponds to the trigger interval and there is no qualified candidate executable window, the dynamic timing queue writes a delayed execution item and the longest waiting boundary. When the joint encoding corresponds to the safe interval, the dynamic timing queue either waits or does not output control commands. The advantage of this embodiment is that it puts the urgency of the dust load and the executableness of the window into the same queue decision structure, avoiding control conflicts caused by their separation.

[0036] Furthermore, the division of the safe zone, approach zone, and trigger zone is updated by the processor based on the current value of the dust load virtual clock and the trigger records in the historical dust removal cycle. Specifically, the processor associates the current value of the dust load virtual clock before dust removal in the historical dust removal cycle with the dust removal recovery record. When the dust removal recovery record indicates that the channel state has recovered differently at a lower virtual clock value, the processor adjusts the lower boundary of the approach zone to the lower virtual clock value. When the dust removal recovery record indicates that the channel state remains stable at a higher virtual clock value, the processor retains the upper boundary of the trigger zone and delays the entry condition of the approach zone. The zone adjustment is constrained by the recovery sample cluster. A single abnormal recovery record is only written to the observation record and does not directly change the zone boundary. The dynamic timed queue rereads the current value of the dust load virtual clock and updates the queue item type after the zone boundary is updated. The advantage of this embodiment is that it avoids the long-term fixed division of the dust load zone and keeps the queue item generation consistent with the actual dust removal recovery state.

[0037] In this embodiment, the processor configures the queue generation time, expected execution time, source continuous time slice, and corresponding current value of the dust load virtual clock for each early execution item, delayed execution item, or immediate execution item in the dynamic timing queue. When a new continuous time slice enters the operating condition feature sequence, the processor recalculates the overlap relationship between the expected execution time and the updated candidate executable window. If the change in the job phase sequence causes the original expected execution time to no longer be within the high-priority candidate executable window, the processor marks the corresponding queue item as a reordering item and rewrites it into the dynamic timing queue according to the latest joint encoding. If the original queue item is an early execution item but the new dust load virtual clock falls back to the safe range or the candidate executable window disappears, the processor converts it into a hold-and-wait item. If the original queue item is a delayed execution item and the new candidate executable window meets the qualification conditions, the processor converts it into an immediate execution item or an early execution item. During the reordering process, the original source continuous time slice and the original current value of the dust load virtual clock are retained so that feedback and traceability can be performed after dust removal is executed. The advantage of this embodiment is that the dynamic timing queue is updated with the job phase change and the source data of the queue item is not lost.

[0038] In an optional embodiment, the queue reordering process includes clearing expired queue items, merging conflicting queue items, and retaining source time slices. Specifically, the processor determines whether multiple queue items in the dynamic timing queue point to similar expected execution times. If multiple queue items originate from consecutive adjacent time slices and correspond to the same dust load virtual clock increment process, the processor retains the queue item with the higher current value of the dust load virtual clock as the main queue item and writes the remaining queue items as source auxiliary records into the main queue item. If multiple queue items originate from different dust removal cycles and the execution feedback has not yet been completed, the processor does not merge them but sets the earlier generated and unconfirmed completed queue items to a pending feedback state to prevent the later queue item from overwriting the previous dust removal control instruction. If a queue item exceeds the longest waiting boundary and there is no qualified candidate executable window, the processor rewrites it as a conservative execution item and records a conservative execution flag after outputting the dust removal control instruction at the control output interface. The advantage of this embodiment is that it enables the dynamic timing queue to maintain a clear priority relationship and backtracking relationship when multiple trigger sources exist simultaneously.

[0039] Furthermore, the processor divides the dust removal recovery records into recovery sample clusters according to the operation phase, the current value of the virtual clock of the dust load before dust removal, and the dust removal continuity record. Specifically, the processor groups the dust removal recovery records with the same or similar operation phase, similar virtual clock interval, and similar dust removal continuity record into the same recovery sample cluster. Within each recovery sample cluster, the processor extracts the median recovery feature and the discrete recovery feature. The median recovery feature is used to represent the normal recovery level under similar dust removal conditions, and the discrete recovery feature is used to represent the fluctuation range of the recovery record under similar dust removal conditions. The processor performs difference calculation between the latest dust removal recovery record and its corresponding recovery sample cluster. When the difference calculation result continuously falls into the decay range, the processor reduces the post-dust removal zeroing ratio of the corresponding operation phase and merges the un-zeroed part into the residual dust load. When the difference calculation result returns to the stable range, the processor restores the historical update step size of the basic dust contribution factor. The advantage of this embodiment is that by distinguishing between single anomalies and continuous decay through recovery sample clusters, the retention of residual dust load and the correction of basic dust contribution factor are supported by similar samples.

[0040] The difference between sample clusters is recovered according to the formula: ; Among them, Q Indicates the first The difference between the dust removal recovery record and its corresponding recovery sample cluster. This indicates the recovered sample cluster to which the m-th dust removal operation belongs. The median recovery characteristics, Indicates the cluster of recovered samples. Discrete recovery characteristics, This represents a stable quantity used to avoid a denominator of zero, such as the first... Secondary dust removal =0.42, belonging to the recovered sample cluster =0.60, =0.12, =0.01, then The processor uses the interval states of multiple consecutive difference values ​​to determine whether the zeroing ratio should be reduced and whether the historical update step size should be restored.

[0041] In this embodiment, the establishment and updating of recovery sample clusters adopts a rolling storage method. The processor retains several recovery sample clusters for each working phase. When a new dust removal recovery record enters, it first determines the candidate recovery sample clusters based on the working phase, and then determines the attribution based on the current value of the virtual clock of dust load before dust removal and the dust removal continuous record. If there is no matching recovery sample cluster, the processor establishes a new recovery sample cluster and marks it as a sample shortage state. Recovery sample clusters in the sample shortage state only participate in record saving and do not participate in zeroing ratio correction. They participate in difference calculation after the number of samples reaches the stable judgment condition. If a recovery sample cluster has not been updated by a new dust removal recovery record for a long time, the processor reduces its priority level for participating in difference calculation, but does not delete the historical record, so that it can still be compared when the loader re-enters the same working conditions. The advantage of this embodiment is that it enables the comparison of dust removal recovery records under similar working conditions and avoids the mixing of data between different dust loads and different working phases.

[0042] Furthermore, after the processor outputs the dust removal control command through the control output interface, it reads the dust removal execution feedback time, feedback continuity record, and execution completion flag, and establishes a write-back relationship with the expected execution time, source continuous time slice, and current value of the dust load virtual clock in the corresponding queue item. Specifically, when the control output interface outputs the dust removal control command, the processor marks the corresponding queue item as output and records the output time. When the dust removal execution feedback arrives, the processor matches the feedback time with the expected execution time, matches the feedback continuity record with the queue item type, and matches the execution completion flag with the dust removal execution record. If the execution completion flag matches the feedback continuity record, the processor resets or partially retains the dust load virtual clock according to the dust removal recovery record. If the execution completion flag does not match the feedback continuity record, the processor pauses the timing reset of the corresponding queue item and writes the current value of the dust load virtual clock into the temporary holding state. When the next continuous time slice confirms the existence of a qualified candidate executable window, the temporary holding state re-participates in the joint encoding. The advantage of this embodiment is that the virtual clock will not be erroneously cleared when the dust removal execution feedback is incomplete or inconsistent.

[0043] In an optional embodiment, the execution feedback write-back also includes state backfilling of the source continuous time slice and queue item closure processing. Specifically, when the execution completion mark is consistent with the feedback continuity record, the processor writes the source continuous time slice, output time, feedback time, dust removal recovery record, and reset ratio into the same closed record, and transfers the corresponding queue item in the dynamic timing queue from the active queue to the historical queue. When the execution completion mark is inconsistent with the feedback continuity record, the processor retains the corresponding queue item in the active queue and marks it as a temporary hold state, while prohibiting new early execution items from overwriting the queue item. If a qualified candidate executable window is confirmed in a subsequent continuous time slice, the processor calculates a new queue item type together with the latest joint encoding. If a qualified candidate executable window is still not confirmed in a subsequent continuous time slice, the processor continues to accumulate the current value of the dust load virtual clock and retains the original source continuous time slice until the feedback is consistent or the conservative execution condition is reached. The advantage of this embodiment is that it distinguishes between unclosed control processes and closed control processes through the active queue and the historical queue, and maintains the correspondence between dust removal triggering, execution feedback, and virtual timing reset.

[0044] In this embodiment, the program instructions executed by the processor also include abnormal data isolation logic. This abnormal data isolation logic shares the same time slice index with phase recognition, dust load virtual clock, and dynamic timing queue. Specifically, when a field in engine speed, driving speed, or hydraulic action state experiences a short-term jump, the processor does not directly change the operation phase label based on that field. Instead, it checks whether adjacent fields form a consistent change relationship. If the engine speed increases briefly but the driving speed range, hydraulic action holding state, and channel resistance change do not form a corresponding change, the processor marks that field as an untrusted sample and uses the stable value of adjacent time slices for phase discrimination. If multiple fields simultaneously show changes consistent with the operation phase transition constraint, the processor writes the change into the phase discrimination vector and participates in the operation phase sequence update. Untrusted samples are still retained in the operating condition feature sequence for feedback tracing, but do not participate in candidate executable window prediction and queue item rearrangement. The advantage of this embodiment is that it isolates abnormal data within a unified time slice structure, avoiding abnormal samples from disrupting the continuity of virtual timing and queue output.

[0045] In an optional embodiment, the program instructions stored in the memory are organized into a data layer, a computation layer, and an output layer. The data layer is used to store the operating condition feature sequence, the operation phase sequence, the dust load virtual clock, the dust removal recovery record, the recovery sample cluster, and the dynamic timing queue. The computation layer is used to perform phase discrimination vector construction, basic dust contribution factor correction, candidate executable window prediction, joint encoding, queue rearrangement, and feedback consistency verification. The output layer is used to output dust removal control instructions to the dust removal execution component through the control output interface and receive dust removal execution feedback. Each record in the data layer has a time slice index or a queue item index. The computation layer only reads data with a valid index and isolated by anomalies. The output layer only responds to queue items in the dynamic timing queue that meet the output conditions. The program instructions are connected through data indexes rather than being directly triggered by fixed time intervals. When the program instructions are executed cyclically in the processor, a closed-loop processing process from running data to dust removal control instructions is formed. The advantage of this embodiment is that it divides data storage, computation decision-making, and output control into clear and implementable logic, which is convenient for deployment in different loader control platforms according to the existing running data access method.

[0046] In this embodiment, after the loader enters the working state, the processor reads existing operating data and updates the working condition characteristic sequence according to time slices. The working phase sequence continues to extend with the time slices. The dust load virtual clock is updated after the time slice corresponding to each confirmed phase ends. The candidate executable window is re-predicted when there is a change at the end of the working phase sequence. The dynamic timing queue is rearranged when the virtual clock interval or window priority value changes. The control output interface outputs a dust removal control command when there is a queue item in the active queue that meets the output conditions. After the dust removal execution feedback arrives, the recovery record calculation and timing reset judgment are entered. If the feedback is consistent and the recovery record is in a stable range... The processor adjusts the residual load coefficient according to the recovery record and enters the next dust removal cycle. If the feedback is inconsistent or the recovery record is within the decay range, the processor retains the residual dust load and makes the temporary holding state re-participate in the joint coding. The cyclic processing does not change the mechanical structure of the loader's dust removal execution component, but changes the timing trigger reference and instruction output timing through computer data processing logic. The advantage of this embodiment is that it forms a continuous technical chain from operation phase recognition, dust load calculation, window matching, queue control to feedback correction, so that the loader's automatic dust removal control can make timing judgments based on the actual operation rhythm and dust load changes.

Claims

1. An intelligent timed automatic dust removal system for loaders, characterized in that, It includes a processor, a memory, and a control output interface. The memory stores program instructions executed by the processor. The processor is configured to access existing operating data of the loader and form a working condition feature sequence by dividing engine speed, driving speed, hydraulic action status, and dust removal execution records into continuous time slices. A work phase sequence is generated based on the described working condition characteristic sequence; The dust load virtual clock is accumulated based on the dust deposition contribution corresponding to each operation phase; A dynamic timing queue is generated when the virtual clock of the dust load matches the predicted executable window, and dust removal control commands are output by the control output interface. When forming the working condition feature sequence, the processor calculates the engine speed change rate, driving speed range, hydraulic action holding state, hydraulic action switching sequence and cumulative running time after dust removal for each continuous time slice, and combines the load lifting relationship, speed direction relationship and hydraulic action sequence relationship between adjacent continuous time slices into a phase discrimination vector. The processor marks the continuous time slice as one of the phases of loading, transporting, lifting, unloading, returning, or idling based on the matching result between the phase discrimination vector and the historical operation cycle segment, and writes it into the operation phase sequence in chronological order. When marking the consecutive time slices, the processor first merges several adjacent consecutive time slices into candidate job segments, and performs stability screening on the phase discrimination vectors within the candidate job segments; When a short-time jump vector exists in the candidate operation segment, the processor performs a consistent reconstruction based on the phase label of the previous candidate operation segment, the phase label of the next candidate operation segment, and the hydraulic action switching order, and writes the reconstructed phase label into the operation phase sequence. When the short-time jump vector is inconsistent with both the preceding and following phase labels, the processor sets the corresponding continuous time slice as a segment to be confirmed and prohibits the segment to be confirmed from participating in the candidate executable window prediction.

2. The intelligent timed automatic dust removal system for loaders according to claim 1, characterized in that, When the processor accumulates the dust load virtual clock, it reads the basic dust contribution factor bound to each phase category according to the phase category in the operation phase sequence, and multiplies or weights the basic dust contribution factor with the engine load level, driving speed level and hydraulic action intensity of the current time slice to obtain the dust contribution amount of the time slice. The processor adds the dust contribution of the time slice to the residual dust load retained in the previous dust removal cycle to generate a non-uniformly increasing virtual timing value, and uses the virtual timing value as the current value of the dust load virtual clock.

3. The intelligent timed automatic dust removal system for loaders according to claim 2, characterized in that, Before generating the dynamic timing queue, the processor predicts candidate executable windows for the next control cycle based on the end phase of the job phase sequence, the phase duration, and the transition relationship between adjacent phases. The processor writes engine load level, hydraulic action conflict level, vehicle speed stability level and dust removal trigger urgency level to the candidate executable windows respectively, and sorts the current value of the dust load virtual clock with the level combination of each candidate executable window to form a timed queue item for immediate execution, early execution, delayed execution or waiting. The control output interface outputs the dust removal control command according to the timed queue item.

4. The intelligent timed automatic dust removal system for loaders according to claim 3, characterized in that, After each dust removal control command is executed, the processor extracts the baseline segment before dust removal and the recovery segment after dust removal, and calculates at least one of the characteristics of air volume change, channel resistance change, or engine load drop, respectively. The processor writes the difference between the post-dust removal recovery segment and the pre-dust removal reference segment into the dust removal recovery record, and corrects the basic dust contribution factor for the same working phase in the next dust removal cycle according to the dust removal recovery record. When the dust recovery record is lower than the dust recovery record level of the same historical operation phase, the residual dust load is retained to the next dust removal cycle according to the difference between the two.

5. The intelligent timed automatic dust removal system for loaders according to claim 4, characterized in that, When forming the timing queue item, the processor establishes a window priority table indexed by the operation phase, engine load level, hydraulic action conflict level, and vehicle speed stability level. The processor divides the current value of the dust load virtual clock into a safe interval, a near interval, and a trigger interval, and jointly encodes the sorting result of the window priority table with the safe interval, the near interval, or the trigger interval. When the joint encoding is a close interval and there is a high-priority candidate executable window, the dynamic timer queue writes an early execution item; When the joint encoding is a trigger interval and there is no qualified candidate executable window, the dynamic timer queue writes the deferred execution item and the longest waiting boundary.

6. The intelligent timed automatic dust removal system for loaders according to claim 5, characterized in that, The processor configures the queue generation time, expected execution time, source continuous time slice, and current value of the corresponding dust load virtual clock for each early execution item, delayed execution item, or immediate execution item in the dynamic timing queue. When a new consecutive time slice enters the operating condition feature sequence, the processor recalculates the overlap relationship between the expected execution time and the updated candidate executable window, and marks queue items whose overlap relationship has changed as items to be rearranged. The items to be rearranged are rewritten into the dynamic time queue according to the joint encoding, while retaining the original continuous time slices.

7. The intelligent timed automatic dust removal system for loaders according to claim 6, characterized in that, The processor divides the dust removal recovery record into recovery sample clusters according to the operation phase, the current value of the virtual clock of dust load before dust removal, and the dust removal continuous record; The processor extracts median recovery features and discrete recovery features within each recovery sample cluster, and performs difference calculation between the latest dust removal recovery record and its corresponding recovery sample cluster. When the difference calculation result continuously falls into the decay range, the processor reduces the dust removal and zeroing ratio of the corresponding operation phase and merges the unzeroed part into the residual dust load. When the difference calculation result returns to the stable range, the processor restores the historical update step size of the basic dust contribution factor.

8. The intelligent timed automatic dust removal system for loaders according to claim 7, characterized in that, After the processor outputs the dust removal control command through the control output interface, it reads the dust removal execution feedback time, feedback continuity record and execution completion flag, and establishes a write-back relationship with the expected execution time, source continuous time slice and current value of the dust load virtual clock in the corresponding queue item. When the execution completion flag is inconsistent with the feedback continuous record, the processor pauses the timing reset of the corresponding queue item and writes the current value of the dust load virtual clock into the temporary hold state; When a qualified candidate executable window is confirmed in the next consecutive time slice, the temporary hold state re-participates in the joint encoding.

Images