A two-stage operational scheduling method and system based on task coverage

Abstract

The application discloses a two-stage operation scheduling method and system based on task coverage, relates to the technical field of intelligent scheduling, and significantly improves the adaptability of the scheduling system to dynamic risks such as cross-day delays through dynamic coverage confidence driving and shadow shift chain pre-embedding technology: multi-source real-time data is fused to generate a coverage confidence label, so that the scheduling scheme automatically adapts to sudden scenes such as position change and meteorological disturbance, and manual intervention is reduced; through the shadow chain pre-generation and automatic switching mechanism, the shift chain is quickly reconstructed at the initial stage of delay triggering, and chain-like resource shortage is avoided; a differentiated resilience coverage threshold ensures the rigid coverage of core tasks, while flexibly adjusting the allocation of auxiliary task resources, so that safety and cost are balanced; the actual rest interval of an employee is dynamically monitored, labor compliance is maintained in combination with shadow chain switching, and operation risk is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent scheduling, in particular to a two-stage operation scheduling method and system based on task coverage. BACKGROUND

[0002] Modern large-scale airport ground service faces the normal challenge of high-frequency cross-day flight delays, especially influenced by factors such as weather and flow control. Preceding flight delays often trigger a domino effect of next-day support resource shortages.

[0003] Under this background, the mainstream two-stage operation scheduling system adopts a task group ring optimization and personnel dynamic assignment architecture: the first stage generates a standardized shift chain based on historical data and flight plans, and the second stage realizes manpower allocation through constraint programming; such a solution optimizes manpower costs and task coverage in a static scenario, but has three dynamic adaptability defects, one of which is the rigid time anchoring defect: the start and end time of the shift preset in the group ring stage is decoupled from the actual execution time of the flight, when cross-day delays cause insufficient actual rest intervals for employees, the system only relies on manual experience to adjust the shift, which easily triggers continuous conflicts in subsequent shifts; the second is the discrete stage decision fragmentation: the group ring stage pursues global cost optimization, but does not pre-embed personnel reuse flexibility such as whether employee A can seamlessly switch to emergency tasks after the early shift delay, resulting in the assignment stage being forced to downgrade to local repair; the third is the limitation of passive response mode: existing dynamic adjustment algorithms such as real-time rolling optimization need to trigger rescheduling after the delay occurs, which cannot predict the delay transmission path and pre-reserve buffer resources, exacerbating the domino effect.

[0004] Some solutions attempt to predict delay impact through reinforcement learning or inject risk hedging strategies such as reserving 5% flexible manpower in the group ring stage to alleviate the problem; however, the former relies on a large amount of labeled data and has high computational delay, making it difficult to meet the airport's minute-level decision-making needs; the latter lacks scenario awareness due to the fixed reservation ratio, resulting in manpower waste in light delays and still being insufficient in severe delays; the latest development proposes a "shadow shift chain" technology, which reduces adjustment overhead by pre-generating backup scheduling paths, but its static pre-generation mode cannot adapt to the randomness and relevance of delay events, and still relies on manual selection of switching time in actual application, failing to form a closed-loop optimization. SUMMARY

[0005] In view of the above existing problems, the present application is proposed.

[0006] The present application provides a two-stage operation scheduling method and system based on task coverage to solve the problem that the traditional two-stage scheduling system is difficult to cope with the domino effect of resource shortages caused by cross-day delays due to rigid time anchoring, stage fragmentation, and passive response mode.

[0007] To solve the above technical problems, the present application provides the following technical solutions:

[0008] In a first aspect, the embodiments of the present application provide a two-stage scheduling method based on task coverage, which comprises,

[0009] Step S1, obtaining flight support task data, employee basic data and dynamic operation data, the dynamic operation data including real-time flight status, gate change information and weather warning signals;

[0010] Step S2, time-space coupling modeling of support tasks based on discretized time window, generating a task demand matrix with coverage confidence label, wherein the coverage confidence is dynamically calculated according to the historical task deviation rate and real-time data fluctuation;

[0011] Step S3, constructing an elastic coverage cost function, the function including a dynamic weight adjustment term, wherein the personnel shortage cost coefficient and the personnel surplus cost coefficient are dynamically updated based on real-time operation situation;

[0012] Step S4, generating a scheduling scheme through two-stage optimization, the first stage generating a coverage-driven shift set based on the task demand matrix and the elastic coverage cost function, and the second stage performing personnel-shift adaptation optimization combined with employee ability label and task pressure coefficient;

[0013] Step S5, pre-generating a shadow shift chain set associated with the main scheduling scheme, and when detecting a cross-day flight delay triggering a domino effect, automatically switching to the optimal shadow shift chain to maintain coverage integrity.

[0014] As a preferred scheme of the two-stage scheduling method based on task coverage, wherein the time-space coupling modeling in step S2 comprises:

[0015] According to the gate coordinate topological relationship, the task transfer time is calculated, and the original task time window is divided into a transfer period and an effective work period;

[0016] Based on the non-uniform time granularity division rule, the transfer loss time compensation is added to the far-gate task to generate a corrected task demand matrix;

[0017] The coverage confidence label is generated by the following method:

[0018] a. Based on the historical same-period flight support task execution data, the task quantity deviation rate of each period is calculated;

[0019] b. Real-time access to gate scheduling system data, when detecting that the gate change rate exceeds the set threshold, the coverage confidence of the associated period is marked with a downgrade;

[0020] c. For the meteorological warning signal coverage period, the confidence value is proportionally reduced according to the warning level.

[0021] As a preferred scheme of the two-stage operation scheduling method based on task coverage, in step S1, the original task demand matrix element is assigned to the coverage confidence C in the following process for each discrete space-time unit i i ;

[0022] Extract historical deviation rate δ from the audited historical task execution system i ,

[0023] Call the flight status interface to obtain the flight delay rate fluctuation ΔS i ,

[0024] Link the gate scheduling system to read the gate change rate ΔG i ,

[0025] Access the meteorological warning platform to obtain the warning level L i ,

[0026] Calculate the real-time fluctuation factor R using historical average value and real-time fluctuation information i , the calculation formula is:

[0027]

[0028] Where a, b, and c are fluctuation sub-weights, ΔS i represents the flight delay rate fluctuation of the i-th period, represents the historical average flight delay rate, ΔG i represents the gate change rate of the i-th period, represents the historical average gate change rate, L i represents the meteorological warning level of the i-th period.

[0029] An exponential decay model is used to fuse historical and real-time information to obtain the initial value C i :

[0030] C i = exp(-(w1δ i +w2R i )),

[0031] Where w1 and w2 are the combined weights of historical deviation rate and real-time fluctuation factor, respectively, δ i is the historical task deviation rate of the i-th period, and R i is the real-time fluctuation factor of the i-th period.

[0032] When the gate change rate exceeds the threshold θ G , the confidence is Ci ←C i ×d G A position change degradation is performed, wherein θ G is a position change rate threshold value, and d G is a position change degradation factor;

[0033] When the weather warning level exceeds the threshold value θ L , the confidence is degraded according to C i ←C i ×d L , wherein θ L is a weather warning level threshold value, and d L is a weather degradation factor;

[0034] The degraded C i is attached to the corresponding position of the original demand matrix to form a final task demand matrix with a coverage confidence label.

[0035] As a preferred scheme of the two-stage operation scheduling method based on task coverage, the elastic coverage cost function satisfies:

[0036] When the real-time flight delay rate exceeds a preset threshold value, the weight of the staff shortage cost coefficient is increased, and the weight of the staff surplus cost coefficient is reduced;

[0037] Differentiated resilience coverage threshold values are set for different task types, rigid coverage constraints are configured for core guarantee task types, and dynamic degraded coverage is allowed for auxiliary task types.

[0038] As a preferred scheme of the two-stage operation scheduling method based on task coverage, the core guarantee task types include aircraft refueling and passenger emergency evacuation passage maintenance.

[0039] The rigid coverage constraint requires that the actual number of on-duty personnel in the task period is greater than or equal to 100% of the demand number, and the personnel qualification matching degree is greater than or equal to 95%.

[0040] As a preferred scheme of the two-stage operation scheduling method based on task coverage, the employee ability label includes a stress resistance coefficient, a multi-task processing score, and a regional movement efficiency, and the task stress coefficient is calculated according to task time consumption, cross-region frequency, and equipment complexity.

[0041] In the second stage optimization, high stress coefficient tasks are preferentially assigned to high stress resistance coefficient employees, and the frequency of continuous execution of high stress tasks by the same employee is limited.

[0042] As a preferred scheme of the two-stage operation scheduling method based on task coverage, the stress resistance coefficient is calculated through the following dimensions:

[0043] Historical high-load task completion rate;

[0044] Emergency task response timeliness;

[0045] Cross-region movement efficiency.

[0046] As a preferred solution of the two-stage scheduling method based on task coverage, the generation of the shadow shift chain set comprises:

[0047] Generating at least three alternative chains for each main shift chain, the alternative chains being obtained by replacing shift start and end times or employee combinations;

[0048] Setting a shadow chain activation condition, when the actual employee rest interval is less than the legal minimum value or the emergency task coverage rate is lower than the resilience threshold, triggering automatic switching;

[0049] The alternative chain generation rule comprises:

[0050] Time offset chain: keeping the original shift task type, shifting the shift start and end times forward or backward by a set time length;

[0051] Personnel replacement chain: replacing a preset proportion of employees in the original shift, preferentially selecting employees with the same qualifications and higher regional movement efficiency;

[0052] Task downgrading chain: allowing a reduction in the number of covered personnel for non-core guarantee tasks to reallocate resources to core tasks.

[0053] As a preferred solution of the two-stage scheduling method based on task coverage, in step S5, at least three alternative chains are generated for each main shift chain, and for any main shift chain C, one alternative chain B is generated in each of the three dimensions of time offset, personnel replacement, and task downgrading k (k = 1, 2, 3), the process comprises:

[0054] Shifting the original start and end times of each shift j in the main chain as a whole to obtain the kth time offset alternative chain:

[0055] ′s,j,k s,j + Δt k e,j,k + Δt e,j k ,

[0056] wherein T s,j , T e,j are the original start and end times of the jth shift, and Δt k ​​​​The time offset of the kth alternative chain is τ, and the value set is {τ, -τ}, τ is a fixed time step;

[0057] The original employee set E of the shift j is replaced by the proportion p j The employee with the optimal combination of moving efficiency and stress resistance is preferentially selected:

[0058] The selection weight of each candidate employee e is defined as:

[0059] W e = λQ e + (1-λ)M e ,

[0060] Wherein, Q e is the stress resistance coefficient of the employee e, M e is the moving efficiency across the area per unit time, and λ∈(0,1) is a weight parameter balancing the two,

[0061] A new set is selected from the available employee pool P:

[0062]

[0063] Wherein, n j is the number of people required for the jth shift, and p is the replacement proportion;

[0064] The coverage requirement r j of the non-core task in the shift j is adjusted by the degradation coefficient β k to generate the kth task degradation alternative chain:

[0065] r j,k = r j × β k ,

[0066] Wherein, r j is the original coverage requirement, and β k ∈(0,1) is the degradation factor of the kth alternative chain;

[0067] After the above three strategies are completed, three alternative chains {B1, B2, B3} are obtained, which can be stored in parallel with the main chain for subsequent automatic switching decision.

[0068] In a second aspect, the present application provides a two-stage operational scheduling system based on task coverage, comprising,

[0069] A data fusion module is used to access flight plan data, real-time operation data and employee ability database, and output a support task flow with time and space labels;

[0070] A coverage calculation engine executes a two-stage operational scheduling method to generate a main scheduling scheme and an associated shadow shift chain.

[0071] a dynamic compensation module, which invokes flexible human resource pool resources and triggers local incremental optimization when detecting a coverage gap or compliance risk;

[0072] a compliance verification interface, which interfaces with a flight execution system in real time, monitors actual rest intervals of employees and task coverage status, and drives shadow chain switching decisions.

[0073] The present application has the beneficial effects that: the present application significantly improves the adaptability of the scheduling system to dynamic risks such as cross-day delays by means of dynamic coverage confidence driving and shadow shift chain pre-embedding technology: fusion of multi-source real-time data to generate coverage confidence labels, so that the scheduling scheme automatically adapts to sudden scenarios such as gate changes and weather disturbances, reducing manual intervention; through the shadow chain pre-generation and automatic switching mechanism, the shift chain is quickly reconstructed at the initial stage of delay triggering, avoiding chain-like resource shortages; differentiated resilience coverage thresholds ensure rigid coverage of core tasks, while flexibly adjusting auxiliary task resource allocation, achieving a balance between safety and cost; dynamically monitoring actual rest intervals of employees, combined with shadow chain switching to maintain labor compliance, reducing operational risks. BRIEF DESCRIPTION OF DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0075] Figure 1 The flowchart of the two-stage operational scheduling method based on task coverage in embodiment 1.

[0076] Figure 2 The framework diagram of the two-stage operational scheduling system based on task coverage in embodiment 1. DETAILED DESCRIPTION

[0077] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings in the specification.

[0078] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0079] Second, the "one embodiment" or "an embodiment" described herein refers to a particular feature, structure, or characteristic included in at least one implementation of the disclosure. The appearances of "in one embodiment" or "in an embodiment" at different places in the specification do not necessarily all refer to the same embodiment, although they can. Furthermore, described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0080] Embodiment 1, Reference Figure 1 and Figure 2 The embodiment provides a two-stage operation scheduling method based on task coverage, comprising the following steps:

[0081] Step S1, obtaining flight support task data, employee basic data and dynamic operation data, the dynamic operation data including real-time flight status, gate change information and weather warning signal;

[0082] Step S2, based on the discretization time window, the space-time coupling modeling of the support task is carried out, and the task demand matrix with coverage confidence label is generated, wherein the coverage confidence is dynamically calculated according to the historical task deviation rate and the real-time data fluctuation;

[0083] The space-time coupling modeling in step S2 includes:

[0084] According to the gate coordinate topological relationship, the task transfer time is calculated, and the original task time window is divided into transfer period and effective operation period;

[0085] Based on the non-uniform time granularity division rule, the transfer loss time compensation is added to the far-gate task, and the corrected task demand matrix is generated;

[0086] The coverage confidence label is generated by the following way:

[0087] a. Based on the historical same period flight support task execution data, the task quantity deviation rate of each period is calculated;

[0088] b. Real-time access to gate scheduling system data, when detecting that the gate change rate exceeds the set threshold, the coverage confidence of the associated period is marked down;

[0089] c. For the weather warning signal coverage period, the confidence value is adjusted by the same ratio according to the warning level;

[0090] In step S1, the way to generate the task demand matrix with coverage confidence label is to assign the original task demand matrix element to the coverage confidence C i for each discrete space-time unit i according to the following process:

[0091] The historical deviation rate δ i is extracted from the audited historical task execution system,

[0092] Call flight status interface to get flight delay rate fluctuation ΔS i ,

[0093] Link the stand scheduling system to read the stand change rate ΔG i ,

[0094] Access the meteorological warning platform to get the warning level L i ,

[0095] Calculate the real-time fluctuation factor R using historical average and real-time fluctuation information i , the calculation formula is:

[0096]

[0097] Where a, b, c are fluctuation sub-weights, ΔS i represents the flight delay rate fluctuation of the i-th period, represents the historical average flight delay rate, ΔG i represents the stand change rate of the i-th period, represents the historical average stand change rate, L i represents the meteorological warning level of the i-th period;

[0098] Use the exponential decay model to fuse historical and real-time information to get the initial value C i :

[0099] C i = exp(-(w1δ i +w2R i )),

[0100] Where w1, w2 are the combined weights of historical deviation rate and real-time fluctuation factor, δ i is the historical task deviation rate of the i-th period, R i is the real-time fluctuation factor of the i-th period;

[0101] When the stand change rate exceeds the threshold θ G , the confidence is degraded according to C i ←C i ×d G , where θ G is the stand change rate threshold, d G is the stand change degradation factor;

[0102] When the meteorological warning level exceeds the threshold θ L , the confidence is degraded according to C i ←C i ×d L , where θ L is the meteorological warning level threshold, d L is the meteorological degradation factor;

[0103] The degraded C i The final task demand matrix with coverage confidence labels is formed by adding the degraded C

[0104] Specifically, this process integrates historical bias and multi-source real-time fluctuations at each spatio-temporal unit level, ensures confidence smooth response through adjustable weight and exponential decay, and makes fine distinction between the impact of gate change and weather warning on coverage capacity through degradation rules, so that the model can adjust in time in the event of sudden changes, and the parameterized design of dynamic factors and degradation factors enhances the model adaptability, which can be flexibly optimized according to the operation strategy;

[0105] Step S3, an elastic coverage cost function is constructed, and the function includes a dynamic weight adjustment term, wherein the personnel shortage cost coefficient and the personnel surplus cost coefficient are dynamically updated based on the real-time operation situation;

[0106] The elastic coverage cost function satisfies:

[0107] When the real-time flight delay rate exceeds the preset threshold, the weight of the personnel shortage cost coefficient is increased, and the weight of the personnel surplus cost coefficient is reduced;

[0108] Differentiated resilience coverage thresholds are set for different task types, and rigid coverage constraints are configured for core support task types, and dynamic degraded coverage is allowed for auxiliary task types;

[0109] The core support task types include aircraft refueling and passenger emergency evacuation passage maintenance;

[0110] The rigid coverage constraint requires that the actual on-duty number in the task period is ≥ 100% of the demand number, and the personnel qualification matching degree is ≥ 95%;

[0111] Step S4, a scheduling scheme is generated through two-stage optimization, the first stage generates a coverage-driven shift set based on the task demand matrix and the elastic coverage cost function, and the second stage performs personnel-shift adaptation optimization in combination with employee ability labels and task stress coefficients;

[0112] The employee ability label includes stress resistance coefficient, multi-task processing score, and regional movement efficiency, and the task stress coefficient is calculated according to task duration, cross-region frequency, and equipment complexity;

[0113] In the second stage optimization, high stress coefficient tasks are preferentially assigned to high stress resistance coefficient employees, and the frequency of continuous execution of high stress tasks by the same employee is limited;

[0114] The stress resistance coefficient is calculated through the following dimensions:

[0115] Historical high-load task completion rate, such as the on-time completion rate of tasks over 8 hours in the past 30 days;

[0116] burst task response timeliness, i.e. the deviation of average arrival time at the work point from the standard time;

[0117] cross-zone movement efficiency, i.e. the number of movements between positions per unit time;

[0118] Step S5, a set of shadow shift chains associated with the main scheduling scheme is pre-generated, and when a cross-day flight delay triggers a domino effect, the optimal shadow shift chain is automatically switched to maintain coverage integrity;

[0119] The generation of the set of shadow shift chains includes:

[0120] At least three alternative chains are generated for each main shift chain, and the alternative chains are obtained by replacing the start and end time of the shift or the employee combination;

[0121] Set the shadow chain activation condition, when the actual employee rest interval is less than the legal minimum value or the burst task coverage rate is lower than the resilience threshold, trigger automatic switching;

[0122] The alternative chain generation rules include:

[0123] Time offset chain: keep the original shift task type, shift the start and end time of the shift forward or backward by a set time length;

[0124] Personnel replacement chain: replace a preset proportion of employees in the original shift, and preferentially select employees with the same qualifications and higher regional movement efficiency;

[0125] Task degradation chain: allow reducing the number of covered personnel for non-core support tasks to reallocate resources to core tasks;

[0126] In step S5, in the process of generating at least three alternative chains for each main shift chain, for any one main shift chain C, one alternative chain B is generated in each of the three dimensions of time offset, personnel replacement and task degradation k (k = 1, 2, 3), the process includes:

[0127] Shift the original start and end time of each shift j in the main chain as a whole to obtain the kth time offset alternative chain:

[0128] T s,j,k = T s,j + Δt k ,T e,j,k = T e,j + Δt k ,

[0129] Where T s,j , T e,j are the original start and end time of the jth shift, Δt kLet be the time offset of the kth candidate chain, with a value set of {τ,-τ}, where τ is a fixed time step;

[0130] Proportionally p in the original employee set E of shift j j When selecting replacement members, priority should be given to employees who possess both excellent mobility and resilience under pressure.

[0131] Define selection weights for each candidate employee e:

[0132] W e =λQ e +(1-λ)M e ,

[0133] Among them, Q e M is the compressive strength coefficient of employee e. e Let be the cross-regional mobility efficiency per unit time, and λ∈(0,1) be the weighting parameter balancing the two.

[0134] Select a new set from the available employee pool P:

[0135]

[0136] Where, n j Let p be the number of people required for the j-th shift, and p be the replacement ratio.

[0137] The coverage requirement r for non-core tasks within shift j j According to the downgrade factor β k Adjustments are made to generate the k-th task degradation alternative chain:

[0138] r j,k =r j ×β k ,

[0139] Where, r j For the original coverage requirement, β k ∈(0,1) is the degradation factor of the k-th candidate chain;

[0140] After implementing the above three strategies, three distinct alternative chains {B1, B2, B3} are obtained, which can be stored in parallel with the main chain for subsequent automatic switching decisions.

[0141] Specifically, through the above three-dimensional alternative chain generation method, the system has multiple solutions when facing cross-day delays or sudden gaps. The time offset strategy can quickly adjust the overall operation window to deal with delayed or advanced flight dynamics. The personnel replacement strategy combines employee stress resistance and mobility efficiency to ensure continuous coverage of key positions with the optimal combination. The task degradation strategy allows limited resources to be concentrated on core tasks when resources are scarce, while maintaining safety and compliance.

[0142] The embodiment also provides a two-stage operation scheduling system based on task coverage, comprising:

[0143] a data fusion module for accessing flight plan data, real-time operation data and employee capacity database, and outputting guarantee task flow with space-time tags;

[0144] a coverage calculation engine for executing a two-stage operation scheduling method, and generating a main scheduling scheme and an associated shadow shift chain;

[0145] a dynamic compensation module for calling flexible human resource pool resources and triggering local incremental optimization when detecting coverage gaps or compliance risks;

[0146] a compliance verification interface for real-time docking with a flight execution system, monitoring actual rest intervals of employees and task coverage states, and driving shadow chain switching decisions.

[0147] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and all should be covered in the scope of the claims of the present application.

Claims

1. A two-stage task-coverage-based operational scheduling method, characterized in that, The application relates to a flight crew scheduling method based on dynamic resilience covering, comprising the following steps: Step S1, obtaining flight guarantee task data, employee basic data and dynamic operation data, wherein the dynamic operation data comprises real-time flight status, gate change information and meteorological warning signals; Step S2, time-space coupling modeling of the guarantee task based on a discretized time window, a task demand matrix with a covering confidence label is generated, wherein the covering confidence is dynamically calculated according to a historical task deviation rate and real-time data fluctuation; Step S3, constructing an elastic covering cost function, wherein the function comprises a dynamic weight adjustment term, wherein the personnel shortage cost coefficient and the personnel surplus cost coefficient are dynamically updated based on the real-time operation situation; Step S4, generating a scheduling scheme through two-stage optimization, the first stage generates a covering-driven shift set based on the task demand matrix and the elastic covering cost function, and the second stage carries out personnel-shift adaptation optimization in combination with the employee ability label and the task pressure coefficient; Step S5, pre-generating a shadow shift chain set associated with the main scheduling scheme, when a cross-day flight delay triggers a domino effect, the optimal shadow shift chain is automatically switched to maintain covering integrity; In step S1, the manner of generating the task demand matrix with coverage confidence labels is that, for each discrete space-time unit , the original task demand matrix element is assigned with a coverage confidence in the following flow ; extracting historical deviation rates from an audited historical mission execution system , Calling a flight status interface to obtain flight delay rate fluctuations , Linking stand scheduling system to read stand change rate , Accessing a weather alert platform to obtain an alert level , Real-time volatility factors are calculated using historical averages and real-time volatility information The formula is: ， wherein a, b, c are fluctuation sub-weights, represents the first period flight delay rate fluctuation, represents the historical average flight delay rate, represents the first period gate change rate, represents the historical average gate change rate, represents the first period weather warning level; An exponential decay model is used to fuse historical and real-time information to obtain an initial value : ， wherein, , are the historical deviation rate and the real-time fluctuation factor combined weight respectively, is the historical task deviation rate of the first time period, is the real-time fluctuation factor of the first time period; When the camera position change rate exceeds a threshold a confidence level is degraded by a camera position change penalty, wherein the camera position change penalty is a function of the camera position change rate, the camera position change rate threshold is a function of the confidence level, the camera position change penalty is a function of the confidence level. when the weather alert level exceeds a threshold value a confidence level is downgraded by a weather downgrade factor, wherein the weather alert level threshold value is a weather alert level threshold value, the weather downgrade factor is a weather downgrade factor; The degraded The final task demand matrix with coverage confidence labels is formed by adding the degraded demand matrix to the original demand matrix at the corresponding positions. The generation mode of the shadow shift chain set comprises: At least three alternative chains are generated for each main shift chain, and the alternative chains are obtained by replacing the start and end times of the shift or the employee combination; The shadow chain activation condition is set, when the actual employee rest interval is less than the legal minimum value or the sudden task covering rate is lower than the resilience threshold, the automatic switching is triggered; The alternative chain generation rule comprises: Time offset chain: the original shift task type is maintained, and the start and end times of the shift are shifted forward or backward by a set time length; Personnel replacement chain: a preset proportion of employees in the original shift are replaced, and employees with the same qualification and higher regional movement efficiency are preferentially selected; Task degradation chain: the number of covered people is allowed to be reduced for non-core guarantee tasks, and resources are allocated to core tasks.

2. A two-stage task-overlay-based scheduling method as claimed in claim 1, characterized in that, The time-space coupling modeling in step S2 comprises: According to the topological relationship of the gate coordinates, the task transfer time is calculated, and the original task time window is divided into a transfer period and an effective operation period; Based on the non-uniform time granularity division rule, the transfer loss time compensation is added to the far-gate task to generate a modified task demand matrix; The covering confidence label is generated in the following way: a. Based on the historical same-period flight guarantee task execution data, the task quantity deviation rate of each period is calculated; b. Real-time access to gate scheduling system data, when it is detected that the gate change rate exceeds the set threshold, the covering confidence of the associated period is marked with a downgrade; c. For the meteorological warning signal covering period, the confidence value is adjusted in proportion to the warning level.

3. A two-stage task-overlay-based scheduling method as claimed in claim 1, characterized in that, The elastic covering cost function satisfies: When the real-time flight delay rate exceeds the preset threshold, the weight of the personnel shortage cost coefficient is increased, and the weight of the personnel surplus cost coefficient is reduced; Differentiated resilience covering thresholds are set for different task types, rigid covering constraints are configured for core guarantee task types, and dynamic degraded covering is allowed for auxiliary task types.

4. A two-stage task-overlay-based scheduling method as claimed in claim 3, characterized in that, The core guarantee task types comprise airplane refueling and passenger emergency evacuation passage maintenance; The rigid covering constraint requires that the actual number of people on duty in the task period is greater than or equal to 100% of the demand number, and the personnel qualification matching degree is greater than or equal to 95%.

5. A two-stage task-overlay-based scheduling method as claimed in claim 1, characterized in that, The employee capability tags include stress resistance coefficient, multitasking score, and regional movement efficiency. The stress resistance coefficient is calculated according to task duration, cross-region frequency, and equipment complexity. In the second stage optimization, high stress coefficient tasks are preferentially assigned to employees with high stress resistance coefficients, and the frequency of consecutive execution of high stress tasks by the same employee is limited.

6. A two-stage task-overlay-based scheduling method as claimed in claim 5, characterized in that, The stress resistance coefficient is calculated by the following dimensions: Historical high-load task completion rate; Emergency task response time; Cross-region movement efficiency.

7. A two-stage task-overlay-based scheduling method as claimed in claim 1, characterized in that, In step S5, in the process of generating at least three candidate chains for each main shift chain, for any one main shift chain , one candidate chain is generated through each of the three dimensions of time offset, personnel replacement and task degradation , wherein k=1, 2, 3, the process includes: Shift the original start and end times of each shift in the main chain by a constant amount to obtain a first time offset alternative chain: time offset alternative chain: ， Wherein, , are the original start and end time of the first class, is the time offset of the first alternative chain, the value set is , is a fixed time step. to scale in shifts of the original staff set replace members, priority is given to employees with high mobility and stress resistance For each candidate employee Define selection weights: ， wherein, the stress coefficient of employees, the efficiency of moving across zones per unit of time, is a weight parameter balancing the two,​ Selecting a new set from the available pool of employees : ， wherein, is the first the number of people required for a shift, is the replacement ratio; Coverage requirements for non-core tasks within a shift are adjusted by a degradation factor to generate a first task degradation candidate chain: ， wherein, is the original coverage requirement, is the first is the degradation factor for the alternative chain. After the above three strategies are completed, three different alternative chains are obtained , which can be stored in parallel with the main chain for subsequent automatic switching decisions.

8. A two-stage operational scheduling system based on task coverage, based on the two-stage operational scheduling method based on task coverage of any one of claims 1-7, characterized in that, The data fusion module is used to access flight plan data, real-time operation data, and employee capability database, and output support tasks with space-time tags; The coverage calculation engine executes a two-stage scheduling method to generate a main scheduling scheme and an associated shadow shift chain; The dynamic compensation module calls flexible human resources and triggers local incremental optimization when a coverage gap or compliance risk is detected; The compliance verification interface is connected to the flight execution system in real time to monitor the actual rest interval of the employee and the task coverage state, and to drive the shadow chain switching decision. ​

Images