An inventory management method and system for the support of rescue helicopter parts

Abstract

This invention relates to the field of aerospace logistics support and intelligent supply chain management technology, specifically to an inventory management method and system for the support of aviation materials for rescue helicopters. The method includes: collecting operation and maintenance data, environmental monitoring data, and equipment attribute data of the target equipment; generating historical consumption feature vectors and multi-dimensional state feature vectors; constructing a nonlinear coupled stress field model and calculating the comprehensive stress index; performing a stress-weighted time axis mapping operation in the time dimension to unify the equipment aging benchmark under different environments; reconstructing demand distribution parameters to generate a dynamic demand distribution model; calculating the dynamic safety stock and reorder point based on the dynamic demand distribution model; and generating tiered procurement instructions and inventory warning signals based on the dynamic safety stock and reorder point. This invention effectively avoids shortages or stockpiles of aviation materials caused by drastic environmental fluctuations and significantly improves the accuracy of support under complex operating conditions.

Description

Technical Field

[0001] This invention relates to the field of aerospace logistics support and intelligent supply chain management technology, specifically to an inventory management method and system for the support of aviation materials for rescue helicopters. Background Technology

[0002] With the continuous evolution of aviation equipment technology and the diversification of deployment environments, the complexity of aviation material support operations has increased significantly. This complexity poses serious challenges to demand forecasting and inventory management, especially in dealing with extreme environmental coupling and sparse fault data.

[0003] Currently, aircraft parts inventory management typically relies on traditional linear statistical forecasting methods. Technicians analyze historical operational data such as flight logs and maintenance records, combining this with simple empirical values ​​to set safety stock and reorder points. While these methods can provide basic support under standard operating conditions, they often fail to reflect the real-time, dynamic impact of external environmental stress on component lifespan. Furthermore, existing statistical forecasting methods have significant limitations when facing complex operating conditions. On one hand, simple linear models cannot effectively quantify the nonlinear coupling relationship between environmental factors such as salt spray, humidity, and temperature differences and equipment age, leading to a high likelihood of both critical component shortages and stockpiles of obsolete materials in extreme environments. On the other hand, when the amount of fault data is small and the sample is sparse, traditional statistical methods are prone to parameter bias, making it difficult to accurately capture long-tail risks. Moreover, procurement orders are easily affected by small data fluctuations, resulting in frequent oscillations and reducing the stability and efficiency of supply chain management. Therefore, accurately quantifying the accelerating effect of multidimensional environmental stress on equipment wear and tear, and achieving precise dynamic inventory optimization under small sample conditions, has become a pressing technical problem in the field of aircraft parts support. Summary of the Invention

[0004] The purpose of this invention is to provide an inventory management method and system for the support of aviation materials for rescue helicopters. Specifically, the technical solution of this invention is as follows:

[0005] An inventory management method for the support of rescue helicopter parts includes the following steps:

[0006] S1. Collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment; clean and time-series align the operation and maintenance data to generate historical consumption feature vectors; perform correlation mapping between environmental monitoring data and equipment attribute data to generate multi-dimensional state feature vectors.

[0007] S2. Construct a nonlinear coupled stress field model, which includes a coupling operator and a sensitivity constant. Input the multidimensional state feature vector into the nonlinear coupled stress field model and solve for the comprehensive stress index.

[0008] S3. Perform stress-weighted time axis mapping operation in the time dimension. Based on the comprehensive stress index, the physical operating time of the target equipment is mapped into the equivalent loss time to unify the equipment aging benchmark under different environments.

[0009] S4. Reconstruct the demand distribution parameters. Based on the equivalent loss time and historical consumption feature vector, calculate the variance inflation correction coefficient. Use the variance inflation correction coefficient to perform nonlinear correction on the standard deviation of the historical demand distribution to generate a dynamic demand distribution model.

[0010] S5. Based on the dynamic demand distribution model, and combined with the preset supply cycle parameters and service level targets, calculate the dynamic safety stock and reorder point.

[0011] S6. Based on the dynamic safety stock level and reorder point, generate tiered procurement instructions and inventory warning signals, and send the procurement instructions to the supply chain management system to trigger resource allocation.

[0012] Optionally, S1 specifically includes:

[0013] S11. Collect the historical fault occurrence time and physical cumulative operating time of the target equipment as operation and maintenance data, which reflects the natural degradation trajectory of the equipment in the physical time dimension; S12. Obtain salt spray concentration data, humidity data and temperature difference change data of the deployment location of the target equipment as environmental monitoring data, which characterizes the corrosive pressure of the external environment on the equipment.

[0014] S12. Extract the current age data and inherent reliability parameters of the target equipment as equipment attribute data;

[0015] S13. Perform sparsity detection on the operation and maintenance data. When the data volume is lower than the preset statistical threshold, it is marked as a sparse sample and the parameter enhancement process is started. The equivalent loss time is used as a bridge to map the failure samples in the heterogeneous environment to the target environment space to achieve sample expansion.

[0016] Optionally, S2 specifically includes:

[0017] S21. Define the calculation logic of the comprehensive stress index, which is determined by the utilization rate coefficient, equipment age coefficient and environmental coefficient.

[0018] S22. Establish a nonlinear coupling relationship between the utilization rate coefficient, equipment age coefficient, and environmental coefficient, and construct a stress function containing interaction terms;

[0019] S23. The interaction term represents the exponential destructive effect of environmental factors on aging equipment. The synergistic destructive effect of environment and age is quantified by multiplying the environmental coefficient with the base of the natural logarithm of the equipment age coefficient.

[0020] S24. Use historical regression analysis to determine the sensitivity constant in the stress function and output the comprehensive stress index that changes dynamically over time.

[0021] Optionally, S3 specifically includes:

[0022] S31. Discretize the physical running time into continuous time elements;

[0023] S32. Within each time micro-element, obtain the current comprehensive stress index;

[0024] S33. Perform integral calculation on the comprehensive stress index within the physical operating time range, stretch or compress each scale on the physical time axis, and output the accumulated equivalent loss time; S34. Use the equivalent loss time to replace the physical operating time as the independent variable, project the equipment failure data under different environmental stresses to the standard stress space, and realize the homogenization processing of heterogeneous data.

[0025] Optionally, S4 specifically includes:

[0026] S41. Calculate the mean and standard deviation of basic demand based on historical consumption feature vectors;

[0027] S42. Calculate the arithmetic square root of the comprehensive stress index, and define the arithmetic square root as the variance expansion correction coefficient, which characterizes the degree to which environmental stress amplifies the uncertainty of the fault.

[0028] S43. Multiply the standard deviation of basic demand by the variance inflation correction factor to obtain the corrected dynamic standard deviation;

[0029] S44. Keep the mean of basic demand unchanged or make linear fine-tuning, and use the corrected dynamic standard deviation to reconstruct the dispersion parameter of the probability distribution function to generate a dynamic demand distribution model that reflects long-tail risk.

[0030] Optionally, S5 specifically includes:

[0031] S51. Obtain the risk level label of the target component, and match the corresponding service level confidence coefficient in the preset decision matrix according to the risk level label;

[0032] S52. Based on the uncertainty of supply cycle parameters and the corrected dispersion of the dynamic demand distribution model, construct a joint variance formula.

[0033] S53. Calculate the safety buffer using the service level confidence coefficient and the joint variance formula, and then combine the predicted demand with the safety buffer to determine the dynamic safety stock.

[0034] S54. Set the Schmitt trigger logic to update the inventory control parameters only when the calculated dynamic safety stock change exceeds the preset dead zone threshold, in order to prevent frequent fluctuations in purchase orders.

[0035] Optionally, S6 specifically includes: real-time monitoring of the accumulation rate of equivalent wear time based on digital twin technology, predicting the remaining time window for the target component to reach the physical failure threshold; when the remaining time window is less than the supply cycle parameter, bypassing the safety stock check logic, directly generating a mandatory pre-positioning purchase order to achieve zero inventory pre-positioning guarantee.

[0036] An inventory management system for the support of aviation materials for rescue helicopters includes the following modules:

[0037] The multidimensional data acquisition module is used to collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment, and generate historical consumption feature vectors and multidimensional status feature vectors.

[0038] The stress field modeling module is used to construct a nonlinear coupled stress field model and calculate the comprehensive stress index based on multidimensional state eigenvectors.

[0039] The time dimension transformation module is used to perform stress-weighted time axis mapping operations, converting physical running time into equivalent loss time.

[0040] The distribution parameter reconstruction module is used to calculate the variance inflation correction coefficient and perform nonlinear correction on the standard deviation of historical demand distribution to generate a dynamic demand distribution model.

[0041] The inventory decision calculation module is used to calculate the dynamic safety stock level and reorder point by combining supply cycle parameters and service level targets.

[0042] The instruction generation and execution module is used to generate tiered procurement instructions and inventory warning signals, and to trigger supply chain resource allocation.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. This invention constructs a nonlinear coupled stress field model to deeply explore the complex synergistic effect between environmental factors, utilization rate and equipment age, solving the problem that traditional linear prediction models cannot accurately quantify the accelerated wear effect of equipment under extreme environments; this mechanism can accurately map the real impact of environmental pressures such as salt spray and humidity on component life, effectively avoiding shortages or stockpiles of aviation materials caused by drastic environmental fluctuations, and significantly improving the accuracy of support under complex working conditions.

[0045] 2. This invention addresses the prediction bias caused by sparse aircraft material failure data and insufficient sample size. The proposed solution introduces a parameter enhancement process based on equivalent loss time. By uniformly mapping heterogeneous samples under different high-stress environments to a standard environmental space, it achieves homogeneous fusion and filling of cross-environment data. Without relying on massive amounts of data, it corrects model bias under small sample conditions and provides reliable data support for inventory optimization of new aircraft models or low-frequency failure parts.

[0046] 3. This invention utilizes a variance inflation correction coefficient to dynamically reconstruct the demand distribution, scientifically quantifying the amplification of external stress on failure uncertainty; by reconstructing the probability distribution through the corrected dynamic standard deviation, the model can automatically identify and capture long-tail risks under harsh service conditions, thereby dynamically adjusting safety stock redundancy; this mechanism achieves a deep balance between cost control and high-confidence assurance, greatly enhancing the robustness of the aviation material support system in dealing with sudden failures.

[0047] 4. This invention introduces Schmitt trigger logic and digital twin monitoring technology. On the one hand, by setting a dead zone threshold to filter noise caused by minor fluctuations in demand, it effectively prevents frequent oscillations in procurement orders and ensures the seriousness and stability of logistics allocation. On the other hand, by monitoring the equivalent loss accumulation rate in real time to predict the failure window, it achieves proactive pre-positioning protection, ensuring accurate replacement of old and new components, fundamentally eliminating downtime risks and optimizing overall operational efficiency. Attached Figure Description

[0048] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0049] Figure 1 This is a flowchart of the method of the present invention;

[0050] Figure 2 This is a structural diagram of the system of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0052] Example 1:

[0053] Please see Figure 1 An inventory management method for the support of rescue helicopter parts includes the following steps:

[0054] S1. Collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment; clean and time-series align the operation and maintenance data to generate historical consumption feature vectors; perform correlation mapping between environmental monitoring data and equipment attribute data to generate multi-dimensional state feature vectors.

[0055] S2. Construct a nonlinear coupled stress field model. The nonlinear coupled stress field model includes a coupling operator and a sensitivity constant. Input the multidimensional state feature vector into the nonlinear coupled stress field model and solve for the comprehensive stress index.

[0056] S3. Perform stress-weighted time axis mapping operation in the time dimension. Based on the comprehensive stress index, the physical operating time of the target equipment is mapped into the equivalent loss time to unify the equipment aging benchmark under different environments.

[0057] S4. Reconstruct the demand distribution parameters. Based on the equivalent loss time and historical consumption feature vector, calculate the variance inflation correction coefficient. Use the variance inflation correction coefficient to perform nonlinear correction on the standard deviation of the historical demand distribution to generate a dynamic demand distribution model.

[0058] S5. Based on the dynamic demand distribution model, and combined with the preset supply cycle parameters and service level targets, calculate the dynamic safety stock and reorder point.

[0059] S6. Based on the dynamic safety stock level and reorder point, generate tiered procurement instructions and inventory warning signals, and send the procurement instructions to the supply chain management system to trigger resource allocation.

[0060] This embodiment aims to solve the problem of simultaneous shortage of critical components and accumulation of obsolete materials caused by the use of simple linear statistical prediction when existing technologies face the coupling of sparse fault data and extreme environments.

[0061] The system performs the steps of collecting and vectorizing multidimensional heterogeneous data. It collects target equipment, such as flight logs and maintenance records of S-76C helicopters at different deployment bases, as well as environmental monitoring data such as salt spray concentration and humidity, through a multi-source heterogeneous data fusion module. After cleaning the noise in the raw data, it generates historical consumption feature vectors and multidimensional state feature vectors containing environmental pressure values ​​and equipment age values.

[0062] A nonlinear coupled stress field model is constructed, which introduces a coupling operator to map the multidimensional state eigenvectors into a dimensionless comprehensive stress exponent. Quantify the acceleration of external conditions on equipment lifespan consumption;

[0063] Perform a stress-weighted time axis mapping operation in the time dimension, and calculate the physical runtime based on the comprehensive stress index. Mapped to equivalent loss time This refers to the time required for the equipment to exhibit the same degree of physical degradation under a standard reference environment.

[0064] Based on this, the demand distribution parameters are reconstructed, and the standard deviation of the demand distribution is corrected using the variance inflation correction coefficient to generate a dynamic demand distribution model that can characterize long-tail risk.

[0065] Calculate dynamic safety stock levels and reorder points, and combine supply cycle parameters (LeadTime) and service level targets, such as an airworthiness assurance rate of 98%, to calculate inventory levels covering a specific confidence interval.

[0066] Generate tiered procurement instructions, convert them into electronic data interchange format, and send them to the supply chain management system (SCM) to trigger logistics allocation.

[0067] Example 2:

[0068] S1 specifically includes:

[0069] S11. Collect the historical fault occurrence time and physical cumulative operating time of the target equipment as operation and maintenance data. The operation and maintenance data reflects the natural degradation trajectory of the equipment in the physical time dimension.

[0070] S12. Obtain salt spray concentration data, humidity data, and temperature difference change data at the target equipment deployment location as environmental monitoring data. The environmental monitoring data characterizes the corrosive pressure of the external environment on the equipment.

[0071] S12. Extract the current age data and inherent reliability parameters of the target equipment as equipment attribute data;

[0072] S13. Perform sparsity detection on the operation and maintenance data. When the data volume is lower than the preset statistical threshold, it is marked as a sparse sample and the parameter enhancement process is started. The equivalent loss time is used as a bridge to map the failure samples in the heterogeneous environment to the target environment space to achieve sample expansion. In the parameter enhancement process, the equivalent loss time is used as a bridge to unify the samples in different environments to the standard environment space, such as the standard temperature and humidity space. In this way, the missing samples in the current environment are filled by cross-environment data, which solves the problem of statistical parameter bias under small sample conditions.

[0073] This embodiment is a further refinement of the data acquisition and sparsity processing steps;

[0074] In the process of digitizing the natural degradation trajectory, the system collects historical fault timestamps accurate to the minute and the total flight hours (TFH) of the equipment up to that moment. This maintenance data is defined as reflecting the linear degradation process of the equipment under the natural passage of physical time.

[0075] In the environmental characterization step of corrosion pressure, the system not only obtains meteorological reports, but also specifically selects salt spray concentration data (mg / m³) that affects the metal corrosion rate, relative humidity (%) that affects the aging of composite materials, and daily temperature difference data that characterizes thermal expansion and contraction stress, and normalizes these data to the [0,1] interval.

[0076] Perform sparsity detection and parameter enhancement; the system determines the number of valid fault records within a preset time window. ; In response to data volume falling below a preset statistical threshold, for example The system labels the dataset as sparse samples; in this case, a parameter augmentation process based on multidimensional state feature vectors is initiated:

[0077] Searching for the same model component in other high-stress environments, the comprehensive stress index is: Failure data below ;

[0078] The formula for calculating the equivalent loss time is: ;

[0079] Calculate the comprehensive stress index of the current target environment by calling preset parameters. ;

[0080] Perform reverse mapping This generates virtual failure data points in the current environment; during the initial running phase of this step, calculations are performed. Required sensitivity constant The system uses historical experience values ​​of the same model of component stored in the system as the initial iteration parameters; after the system completes the parameter regression analysis of the current deployment location through step S2, the updated sensitivity constant is used to perform a second correction on the virtual failure data points; through the above steps, external high-value data is projected onto the current sparse space to construct a statistically significant sample space.

[0081] Example 3:

[0082] S2 specifically includes:

[0083] S21. Define the calculation logic of the comprehensive stress index. The comprehensive stress index is jointly determined by the utilization rate coefficient, equipment age coefficient, and environmental coefficient.

[0084] S22. Establish a nonlinear coupling relationship between the utilization rate coefficient, equipment age coefficient, and environmental coefficient, and construct a stress function containing interaction terms;

[0085] S23. The interaction term characterizes the exponential destructive effect of environmental factors on aging equipment. It quantifies the synergistic destructive effect of environment and equipment age by multiplying the environmental coefficient with the base of the natural logarithm, which is exponentially based on the equipment age coefficient.

[0086] S24. Use historical regression analysis to determine the sensitivity constant in the stress function and output the comprehensive stress index that changes dynamically over time.

[0087] This embodiment details the nonlinear calculation logic of the comprehensive stress index; to strictly conform to the definition of the interaction term in the embodiment, the system constructs a nonlinear coupled stress field model in the following form:

[0088] Formula 1: ;

[0089] The parameters are defined as follows:

[0090] : Comprehensive stress index, output variable;

[0091] The source is the system clock, and its physical meaning is the current time.

[0092] Utilization rate coefficient, obtained by normalizing flight frequency;

[0093] The environmental coefficient is calculated by weighting salt spray, humidity, and temperature difference, and corresponds to the environmental coefficient in the example. The specific weight allocation logic is as follows: ,in These are the normalized real-time salt spray concentration, humidity, and daily temperature difference monitoring values, respectively. The preset sensitivity weighting coefficients are used, and the sum of the weights is 1; for structural components with a high proportion of metal material, the weighting coefficients are preferably set to [value missing]. ;

[0094] : Equipment age coefficient, corresponding to the equipment age coefficient in the embodiment;

[0095] The term originates from a mathematical constant and its physical meaning is the base of the natural logarithm.

[0096] : The base of the natural logarithm, with the equipment age coefficient as the exponent, in the corresponding embodiment;

[0097] The mathematical meaning of the interaction item corresponding to the embodiment is: as the machine age increases... As the environmental factor increases, the physical vulnerability of the equipment rises exponentially. As a multiplicative factor, it amplifies the destructive effect on this vulnerability;

[0098] The sensitivity constant was determined through regression analysis of historical failure data using the maximum likelihood estimation (MLE) method; among which, The utilization sensitivity coefficient The environmental coupling strength coefficient. For model constants, The time decay constant is represented by the reciprocal of physical time, ensuring the exponential term in the nonlinear connection field model. The value is dimensionless; the maximum likelihood estimation method minimizes the likelihood function with respect to the moment of equipment failure and uses Newton's iteration method to solve for the parameter combination that maximizes the probability of the observed sample.

[0099] Example 4:

[0100] S3 specifically includes:

[0101] S31. Discretize the physical running time into continuous time elements;

[0102] S32. Within each time micro-element, obtain the current comprehensive stress index;

[0103] S33. Perform integral calculation on the comprehensive stress index within the physical running time range, stretch or compress each scale on the physical time axis, and output the accumulated equivalent loss time.

[0104] S34. By using equivalent loss time instead of physical running time as the independent variable, equipment failure data under different environmental stresses are projected onto the standard stress space to achieve homogenization of heterogeneous data.

[0105] This embodiment details the calculus solution for time dimension transformation;

[0106] The system performs a discretization operation, discretizing the physical runtime into continuous time infinitesimals. ;

[0107] Obtain the current instantaneous comprehensive stress index within each time microelement. ;

[0108] Perform integral calculations and calculate the equivalent loss time using the following formula:

[0109] ;

[0110] The source is calculated, and the physical meaning is the equivalent wear time of the equipment under standard reference environment, with the unit being standard hours;

[0111] The source is a flight recorder; the physical meaning is the physical cumulative running time, in hours.

[0112] The source is a system preset, and its physical meaning is the calculation step size of the discretization.

[0113] The source is the stress field model output, and its physical meaning is the instantaneous comprehensive stress index within the i-th micro-element;

[0114] The source is the discretization process, and its physical meaning is the total number of time-varying infinitesimals;

[0115] The source is a mathematical definition, and its physical meaning is an integral variable, representing continuous physical time;

[0116] System utilization Alternative As an independent variable, equipment failure data under different environmental stresses are projected onto the standard stress space;

[0117] This embodiment uses integral transformation to stretch or compress each scale on the physical time axis according to real-time stress, successfully mapping the heterogeneous fault data that was originally randomly distributed in the physical time dimension to the equivalent loss time axis. This transformation makes the originally discrete fault points exhibit high consistency and regularity after projection, such as being concentrated in specific standard hours, which greatly improves the accuracy of predicting the remaining service life of equipment under different service environments and realizes the homogenization of heterogeneous data.

[0118] Example 5:

[0119] S4 specifically includes:

[0120] S41. Calculate the mean and standard deviation of basic demand based on historical consumption feature vectors;

[0121] S42. Calculate the arithmetic square root of the comprehensive stress index, and define the arithmetic square root as the variance expansion correction factor. The variance expansion correction factor characterizes the degree to which environmental stress amplifies the uncertainty of the fault.

[0122] S43. Multiply the standard deviation of basic demand by the variance inflation correction factor to obtain the corrected dynamic standard deviation;

[0123] S44. Keep the mean of basic demand unchanged or make linear fine-tuning, and use the corrected dynamic standard deviation to reconstruct the dispersion parameter of the probability distribution function to generate a dynamic demand distribution model that reflects long-tail risk.

[0124] This embodiment details how to reconstruct the uncertainty parameters of the demand distribution through variance inflation;

[0125] The system calculates the average basic demand based on historical consumption feature vectors. Standard deviation of basic demand ;

[0126] Calculate the average comprehensive stress index within the statistical period. The arithmetic square root is defined as the variance inflation correction factor. The physical essence of this coefficient is to stretch the basic distribution by the comprehensive inflation index, and map the uncertainty caused by environmental error into a long-tail feature of demand probability density, so as to cover the sudden risks under the power grid operating conditions by increasing the dispersion without increasing the mean.

[0127] Using formula Calculate the corrected dynamic standard deviation;

[0128] Perform mean fine-tuning and distribution reconstruction in step S44:

[0129] Mean linear fine-tuning: Set the fine-tuning coefficient , The value range is set at Between, calculation For mechanical friction components whose failure rate is severely affected by the environment A value of 0.15 is used for sealed electronic components. The value is 0.05;

[0130] Distribution reconstruction: using corrected parameters Constructing a dynamic normal distribution model ;because quilt With significant scaling, the probability density function of the distribution model becomes flatter, and the tail probability increases significantly, thus mathematically quantifying the increase in long-tail risk in adverse environments.

[0131] Example 6:

[0132] S5 specifically includes:

[0133] S51. Obtain the risk level label of the target component, and match the corresponding service level confidence coefficient in the preset decision matrix according to the risk level label;

[0134] S52. Based on the uncertainty of supply cycle parameters and the corrected dispersion of the dynamic demand distribution model, construct a joint variance formula.

[0135] S53. Calculate the safety buffer using the service level confidence coefficient and the joint variance formula, and then combine the predicted demand with the safety buffer to determine the dynamic safety stock.

[0136] S54. Set the Schmitt trigger logic to update the inventory control parameters only when the calculated dynamic safety stock change exceeds the preset dead zone threshold, in order to prevent frequent fluctuations in purchase orders.

[0137] This embodiment involves a specific dynamic safety stock decision-making algorithm;

[0138] The system obtains the risk level label of the target component and matches the corresponding service level confidence coefficient in the decision matrix. ;

[0139] Construct a joint variance formula and use the following formula to calculate the dynamic safety stock:

[0140] Formula 2: ;

[0141] The source is calculated, and the physical meaning is dynamic safety stock.

[0142] The source is decision matrix matching, and the physical meaning is the service level confidence coefficient;

[0143] L: Sourced from historical supply chain records, its physical meaning is the ratio of the average supply cycle to the statistical time unit of basic demand, a dimensionless multiple used to ensure the first term within the square root. and the second term The dimensions of both are the square of the quantity;

[0144] The source is the output of step S4, and its physical implication is the standard deviation of the dynamic demand in the later stage. Its unit of measurement is consistent with the unit of quantity required for aerospace materials, ensuring that the first term within the square root is accurate. This represents the cumulative increase in demand and secondary terms during the supply cycle. The dimensions of both are the square of the quantity;

[0145] The source is the basic forecast, and its physical meaning is the basic forecast demand rate per unit time.

[0146] The source is statistical analysis, and the physical meaning is the standard deviation of the supply cycle;

[0147] The system executes Schmitt trigger logic; in response to the calculated change in dynamic safety stock. Exceeding the preset dead zone threshold Specify dead zone threshold Employing non-precise mapping logic: For a type of aircraft material with extremely high inventory holding costs, set... To maintain a high level of environmental awareness, the calculated value is set at 5%; for low-value, easily consumable Category C aviation materials, the following setting is made: To filter out logistics scheduling disturbances caused by small-scale random fluctuations, the calculated value is set to 20%; for example, 10% of the current inventory. The system then updates the inventory control parameters; otherwise, the current parameters remain unchanged. Dead zone threshold. The setting is based on the A / B / C classification attributes of the components. For high-value, long-cycle Class A aircraft materials, Set to a lower value to maintain high sensitivity; for general consumable Category C aircraft materials, Set to a higher value to reduce redundant fluctuations in logistics instructions;

[0148] This embodiment introduces Schmitt trigger logic to add a low-pass filter to the inventory decision system, effectively filtering out high-frequency noise caused by small fluctuations in dynamic standard deviation. This mechanism prevents frequent jumps between recommended purchases and recommended postponements in purchasing orders, ensuring the stability and seriousness of supply chain orders. At the same time, combined with the joint variance formula, it achieves a dynamic balance between cost and risk.

[0149] Example 7:

[0150] S6 specifically includes:

[0151] Based on digital twin technology, the system monitors the accumulation rate of equivalent wear time in real time and predicts the remaining time window for the target component to reach the physical failure threshold. When the remaining time window is less than the supply cycle parameter, the system bypasses the safety stock check logic and directly generates a mandatory pre-positioned purchase order to achieve zero-inventory pre-positioning guarantee.

[0152] This embodiment details the zero-inventory provisioning logic based on prediction;

[0153] The system uses digital twin technology to monitor the accumulation rate of equivalent time loss in real time;

[0154] Calculate the remaining time window using the following formula:

[0155] ;

[0156] The source is calculated, and its physical meaning is the predicted remaining time before reaching the failure threshold;

[0157] The source is the component manual, and the physical meaning is the physical failure threshold based on material properties;

[0158] The source is real-time calculation, and the physical meaning is up to the current physical moment. The cumulative equivalent time lost;

[0159] The source is the current physical time. The stress field model is output in real time and numerically satisfies... That is, this parameter is directly assigned by the instantaneous comprehensive stress index calculated by the nonlinear connection field model, which represents the equivalent loss time converted per unit physical running time;

[0160] The system executes logical judgments and responds within the remaining time window. Less than the supply cycle parameter If the system determines that there is a risk of shutdown, it will bypass the normal safety stock check logic and directly generate a mandatory pre-purchase order.

[0161] This embodiment establishes a preemptive procurement mechanism through real-time monitoring and reverse engineering. This mechanism ensures that new components arrive precisely at the moment when old components reach the physical failure threshold, achieving theoretical zero-inventory pre-positioning protection. This not only minimizes inventory holding costs but also completely eliminates the risk of aircraft grounding due to logistics delays, achieving the optimization of airworthiness assurance efficiency.

[0162] Example 8:

[0163] Please see Figure 2 An inventory management system for the support of aviation materials for rescue helicopters includes the following modules:

[0164] The multidimensional data acquisition module is used to collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment, and generate historical consumption feature vectors and multidimensional status feature vectors.

[0165] The stress field modeling module is used to construct a nonlinear coupled stress field model and calculate the comprehensive stress index based on multidimensional state eigenvectors.

[0166] The time dimension transformation module is used to perform stress-weighted time axis mapping operations, converting physical running time into equivalent loss time.

[0167] The distribution parameter reconstruction module is used to calculate the variance inflation correction coefficient and perform nonlinear correction on the standard deviation of historical demand distribution to generate a dynamic demand distribution model.

[0168] The inventory decision calculation module is used to calculate the dynamic safety stock level and reorder point by combining supply cycle parameters and service level targets.

[0169] The instruction generation and execution module is used to generate tiered procurement instructions and inventory warning signals, and to trigger supply chain resource allocation.

[0170] This embodiment provides an inventory management system for the support of rescue helicopter parts. This system is used to execute the methods described in any one of embodiments 1 to 7 above. The system's hardware architecture relies on distributed servers and edge computing nodes, and its functional modules specifically include:

[0171] The multidimensional data acquisition module is configured to connect the airborne health management system and the meteorological database to collect the operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment, and perform data cleaning to generate historical consumption feature vectors and multidimensional state feature vectors.

[0172] The stress field modeling module is configured to load nonlinear coupling operators to construct a nonlinear coupled stress field model and to solve the comprehensive stress index in real time based on multidimensional state feature vectors.

[0173] The time dimension transformation module is configured to execute the core stress-weighted time axis mapping algorithm, which is used to map physical running time into equivalent loss time, so as to achieve homogenization and unification of equipment aging benchmarks in heterogeneous environments.

[0174] The distribution parameter reconstruction module is configured to be based on statistical principles to calculate the variance inflation correction coefficient and perform nonlinear correction on the standard deviation of historical demand distribution to generate a dynamic demand distribution model.

[0175] The inventory decision calculation module is configured to load supply chain parameters and combine supply cycle parameters with service level targets to calculate dynamic safety stock and reorder point.

[0176] The instruction generation and execution module is configured to connect to the Enterprise Resource Planning (ERP) system to generate tiered procurement instructions and inventory warning signals, and to trigger supply chain resource allocation.

[0177] This embodiment transforms a complex physical-statistical coupling model into an executable automated process through a modular system architecture; the data flow between modules realizes closed-loop control from physical world perception to supply chain decision execution, ensuring that the aviation material support system can achieve optimal inventory allocation in a data-driven manner under a changing external environment.

[0178] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An inventory management method for the support of aviation materials for rescue helicopters, characterized in that, Includes the following steps: S1. Collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment; clean and time-series align the operation and maintenance data to generate historical consumption feature vectors; perform correlation mapping between environmental monitoring data and equipment attribute data to generate multi-dimensional state feature vectors. S2. Construct a nonlinear coupled stress field model, which includes a coupling operator and a sensitivity constant. Input the multidimensional state feature vector into the nonlinear coupled stress field model and solve for the comprehensive stress index; including: S21. Define the calculation logic of the comprehensive stress index, which is determined by the utilization rate coefficient, equipment age coefficient and environmental coefficient. S22. Establish a nonlinear coupling relationship between the utilization rate coefficient, equipment age coefficient, and environmental coefficient, and construct a stress function containing interaction terms; S23. The interaction term represents the exponential destructive effect of environmental factors on aging equipment. The synergistic destructive effect of environment and age is quantified by multiplying the environmental coefficient with the base of the natural logarithm of the equipment age coefficient. S24. Use historical regression analysis to determine the sensitivity constant in the stress function and output the comprehensive stress index that changes dynamically over time. S3. Perform stress-weighted time axis mapping operation in the time dimension. Based on the comprehensive stress index, convert the physical operating time of the target equipment into equivalent wear time to unify the equipment aging benchmark under different environments; including: S31. Discretize the physical running time into continuous time elements; S32. Within each time micro-element, obtain the current comprehensive stress index; S33. Perform an integral calculation on the comprehensive stress index within the physical operating time range. The integral calculation result is the equivalent loss time of the target equipment. The equivalent loss time is the time required for the equipment to produce the same degree of physical degradation under the standard reference environment. S34. Using equivalent loss time instead of physical running time as the independent variable, equipment failure data under different environmental stresses are projected onto the standard stress space to achieve homogenization of heterogeneous data. S4. Reconstruct the demand distribution parameters. Based on the equivalent loss time and historical consumption feature vector, calculate the variance inflation correction coefficient. Use the variance inflation correction coefficient to perform nonlinear correction on the standard deviation of the historical demand distribution to generate a dynamic demand distribution model. S5. Based on the dynamic demand distribution model, and combined with the preset supply cycle parameters and service level targets, calculate the dynamic safety stock and reorder point. S6. Based on the dynamic safety stock level and reorder point, generate tiered procurement instructions and inventory warning signals, and send the procurement instructions to the supply chain management system to trigger resource allocation.

2. The inventory management method for the support of rescue helicopter parts according to claim 1, characterized in that, S1 specifically includes: S11. Collect the historical fault occurrence time and physical cumulative operating time of the target device as operation and maintenance data. The operation and maintenance data reflects the natural degradation trajectory of the device in the physical time dimension. S12. Obtain salt spray concentration data, humidity data, and temperature difference change data at the deployment location of the target equipment as environmental monitoring data, wherein the environmental monitoring data characterizes the corrosive pressure of the external environment on the equipment; S13. Extract the current age data and inherent reliability parameters of the target equipment as equipment attribute data; S14. Perform sparsity detection on the operation and maintenance data. When the data volume is lower than the preset statistical threshold, it is marked as a sparse sample and the parameter enhancement process is started. The equivalent loss time is used as a bridge to unify the samples in different environments into the standard environment space in order to achieve sample expansion.

3. The inventory management method for the support of rescue helicopter parts according to claim 1, characterized in that, S4 specifically includes: S41. Calculate the mean and standard deviation of basic demand based on historical consumption feature vectors; S42. Calculate the arithmetic square root of the comprehensive stress index, and define the arithmetic square root as the variance expansion correction coefficient, which characterizes the degree to which environmental stress amplifies the uncertainty of the fault. S43. Multiply the standard deviation of basic demand by the variance inflation correction factor to obtain the corrected dynamic standard deviation; S44. Keep the mean of basic demand unchanged or make linear fine-tuning, and use the corrected dynamic standard deviation to reconstruct the dispersion parameter of the probability distribution function to generate a dynamic demand distribution model that reflects long-tail risk.

4. The inventory management method for the support of rescue helicopter parts according to claim 3, characterized in that, S5 specifically includes: S51. Obtain the risk level label of the target component, and match the corresponding service level confidence coefficient in the preset decision matrix according to the risk level label; S52. Based on the uncertainty of the supply cycle parameter and the corrected dispersion of the dynamic demand distribution model, a joint variance formula is constructed; the uncertainty of the supply cycle parameter specifically refers to the average supply cycle and the standard deviation of the supply cycle. S53. Calculate the safety buffer using the service level confidence coefficient and the joint variance formula, and then combine the predicted demand with the safety buffer to determine the dynamic safety stock. S54. Set the Schmitt trigger logic to update the inventory control parameters only when the calculated dynamic safety stock change exceeds the preset dead zone threshold.

5. The inventory management method for the support of rescue helicopter parts according to claim 4, characterized in that, Specifically, S6 includes: real-time monitoring of the accumulation rate of equivalent wear time based on digital twin technology, predicting the remaining time window for the target component to reach the physical failure threshold; when the remaining time window is less than the supply cycle parameter, bypassing the safety stock check logic, directly generating a mandatory pre-positioned purchase instruction to achieve zero inventory pre-positioning guarantee.

6. An inventory management system for the support of rescue helicopter parts, applied to the method described in any one of claims 1 to 5, characterized in that, Includes the following modules: The multidimensional data acquisition module is used to collect operation and maintenance data, environmental monitoring data and equipment attribute data of the target equipment, and generate historical consumption feature vectors and multidimensional status feature vectors. The stress field modeling module is used to construct a nonlinear coupled stress field model and calculate the comprehensive stress index based on multidimensional state eigenvectors. The time dimension transformation module is used to perform stress-weighted time axis mapping operations, converting physical running time into equivalent loss time. The distribution parameter reconstruction module is used to calculate the variance inflation correction coefficient and perform nonlinear correction on the standard deviation of historical demand distribution to generate a dynamic demand distribution model. The inventory decision calculation module is used to calculate the dynamic safety stock level and reorder point by combining supply cycle parameters and service level targets. The instruction generation and execution module is used to generate tiered procurement instructions and inventory warning signals, and to trigger supply chain resource allocation.

Images