Full motion simulator projector life prediction method, system, and device

Abstract

The present application belongs to the field of machine life prediction, and particularly relates to a full-motion simulator projector life prediction method, system and device. It aims to solve the problem of ignoring early signals and static acquisition strategy in the prior art. The present application starts the dark field mode and divides the area to collect multi-dimensional operating parameters; calculates the brightness, color change rate and transmission delay deviation rate; dynamically adjusts the abnormal threshold according to the device working time, historical fluctuation and environmental stability; adaptively adjusts the key area acquisition frequency through multi-condition joint trigger judgment; after extracting the regional and global degradation characteristics, inputting into the LSTM neural network model with attention mechanism, high-precision residual life prediction is realized. The present application effectively overcomes the defects of ignoring early aging signals and static acquisition strategy in the prior art, and significantly improves the early fault identification ability and prediction accuracy.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of machine life prediction, and particularly relates to a full-motion simulator projector life prediction method, system and device. BACKGROUND

[0002] As high-end training equipment, the stability and reliability of the projection system of the full-motion simulator directly affect the training effect and safety. As the core display component of the system, the performance of the projector will gradually degrade with the increase of working time, and accurate prediction of its remaining life is crucial for formulating predictive maintenance strategies and avoiding training interruption.

[0003] Currently, the existing technical solutions in the field of projector life prediction mainly focus on two categories: traditional statistical data-driven methods and random process model-based methods. The traditional statistical data-driven method statistically analyzes the historical failure time data of the equipment, constructs a life distribution function, and then calculates the conditional remaining life of the equipment at a specific time. This method relies heavily on a large amount of complete failure data accumulation and cannot reflect individual differences and real-time operating conditions. The random process model-based method is based on probability and statistics theory, models the evolution law of performance degradation variables, and can give the prediction result of remaining life in the form of probability distribution, not only providing point estimation, but also describing the uncertainty of prediction. However, this method still faces significant challenges in practical application: first, the performance degradation variables are mostly single macroscopic parameters such as brightness and chrominance, which are difficult to extract deep features directly related to early aging; second, the model construction relies on single variable degradation process and fails to fully exploit the coupling information between multiple parameters (such as electrical, optical and thermal parameters), limiting the improvement of prediction accuracy.

[0004] In summary, the existing technology has the following main defects when applied to the life prediction of the full-motion simulator projector: first, the data collection lacks pertinence and generally ignores the capture of early aging signals (such as weak brightness anomalies and color drift) in dark field mode, making it difficult to identify early failures in time; second, it fails to effectively utilize key performance indicators (such as transmission delay response time) closely related to optical loss and circuit aging, and fails to combine them with standardized test specifications (such as QTG test), resulting in incomplete input features of the prediction model; third, the static and rigid collection strategy lacks sufficient parameter collection frequency in areas with rapid changes, making it difficult to capture the dynamic degradation process of key parameters, and the fixed threshold mechanism is difficult to adapt to individual differences and environmental changes, leading to false positives or false negatives; fourth, the prediction model is not sensitive to local anomalies and fails to distinguish the importance of different regional features, making it difficult to focus on abnormal changes in high-impact areas, resulting in large prediction errors.

[0005] Based on this, the application provides a full-motion simulator projector life prediction method, system and device. SUMMARY

[0006] In order to solve the above problems in the prior art, that is, the prior art ignores early signals and the collection strategy is static, the application provides a full-motion simulator projector life prediction method, system and device.

[0007] In a first aspect of the application, a full-motion simulator projector life prediction method is provided, which comprises:

[0008] Starting the dark field mode of the projector, dividing the projection picture into multiple independent regions, and collecting the running state parameters, transmission delay response time and environmental temperature and humidity parameters of each region;

[0009] According to the running state parameters of each region, the brightness change rate and the color change rate are calculated, and the current transmission delay response time is compared with the standard reference value to obtain the transmission delay deviation rate;

[0010] According to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability, the abnormal judgment threshold is dynamically adjusted;

[0011] According to the brightness change rate, the color change rate and the transmission delay deviation rate and the adjusted abnormal judgment threshold, it is judged whether the cross-parameter joint triggering condition is met; the joint triggering condition includes double-parameter correlation triggering, transmission delay and physical parameter coupling triggering or multi-parameter trend consistency triggering;

[0012] The collection frequency of the running state parameters of the region meeting the triggering condition is increased and the delay sampling interval is shortened, and the corresponding collection frequency and interval of the region not meeting the condition are reduced;

[0013] The running state parameters, transmission delay response time and environmental temperature and humidity parameters collected in real time are preprocessed, and the region brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature are extracted;

[0014] The feature parameters are input into a long short-term memory neural network model with attention mechanism, and the remaining life prediction value is output.

[0015] Further, according to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability, the abnormal judgment threshold is dynamically adjusted, and the method is:

[0016] According to the different aging stages to which the cumulative working time of the device belongs, a corresponding first correction coefficient is selected, wherein the aging stage is divided according to the working time range;

[0017] calculating a second correction coefficient based on a difference between the historical parameter fluctuation amplitude and the device average fluctuation standard deviation;

[0018] determining a third correction coefficient based on whether the environmental temperature and humidity fluctuation data exceeds a stability threshold value;

[0019] multiplying the initial threshold value by the first correction coefficient, the second correction coefficient and the third correction coefficient to obtain a final abnormality judgment threshold value after dynamic adjustment.

[0020] Further, the difference between the historical parameter fluctuation amplitude and the device average fluctuation standard deviation is subjected to ratio operation with the device average fluctuation standard deviation, and the operation result is mapped into a preset coefficient range to obtain the second correction coefficient.

[0021] Further, according to the brightness change rate, the color change rate, the transmission delay deviation rate and the adjusted abnormality judgment threshold value, it is judged whether the cross-parameter joint triggering condition is met, and the method is:

[0022] The double-parameter correlation triggering condition is that the change rate or the deviation rate of any two parameters simultaneously reaches a set proportion of the respective abnormality judgment threshold value.

[0023] The transmission delay and physical parameter coupling triggering condition is that the transmission delay deviation rate reaches a set proportion of the abnormality judgment threshold value, and the current change rate or the temperature change rate of the corresponding area exceeds the respective set fluctuation limit value.

[0024] The multi-parameter trend consistency triggering condition is that in a plurality of continuous acquisition cycles, the brightness change rate, the color change rate and the transmission delay deviation rate all show a monotonic increasing trend, and the change rate of at least two parameters exceeds a set proportion of the abnormality judgment threshold value.

[0025] Further, the real-time collected running state parameters, transmission delay response time and environmental temperature and humidity parameters are preprocessed, and the regional brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature are extracted, and the method is:

[0026] The real-time collected running state parameters, transmission delay response time and environmental temperature and humidity parameters are preprocessed, and the preprocessing includes identifying abnormal values, correcting abnormal values and filtering processing;

[0027] Based on the preprocessed data, regional feature parameters and global feature parameters are extracted, wherein:

[0028] The regional feature parameters are obtained by calculating the average attenuation amount of the brightness value of each region, the average value of the color change rate and the average value of the transmission delay deviation rate.

[0029] The parameter correlation degree feature is obtained by calculating the correlation coefficient of the transmission delay deviation rate and the current change rate and the correlation coefficient of the temperature change rate.

[0030] The global fluctuation entropy feature is obtained by calculating the sample entropy of the luminance change rate, the color change rate and the delay deviation rate sequence of all regions.

[0031] Further, the feature parameters are input into a long short-term memory neural network model with an attention mechanism, and a remaining life prediction value is output, and the method is:

[0032] The regional feature parameters and the global feature parameters are combined into a model input vector;

[0033] The input vector is input into a pre-trained long short-term memory neural network model;

[0034] The long short-term memory neural network model calculates the weight of each regional feature parameter through the attention mechanism layer inside it, and performs weighted summation on the regional feature parameters to focus on the features of the high-impact region;

[0035] The processing layer of the long short-term memory neural network model operates on the weighted features and outputs a prediction value representing the normalized remaining life;

[0036] The normalized remaining life prediction value is denormalized to obtain the final remaining life prediction value.

[0037] Further, the operating state parameters at least include luminance parameters, color parameters, current parameters and temperature parameters.

[0038] Further, when the prediction value is lower than a preset threshold, a maintenance warning is issued.

[0039] In a second aspect, the application provides a full-motion simulator projector life prediction system for a full-motion simulator projector life prediction method, the system comprising:

[0040] A data acquisition module configured to start the dark field mode of the projector, divide the projection picture into multiple independent regions, and collect operating state parameters, transmission delay response time and environmental temperature and humidity parameters of each region;

[0041] A data calculation module configured to calculate the luminance change rate and the color change rate according to the operating state parameters of each region, and obtain the transmission delay deviation rate by comparing the current transmission delay response time with a standard reference value;

[0042] A threshold updating module configured to dynamically adjust the abnormality determination threshold according to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability;

[0043] a trigger condition judging module configured to judge whether a cross-parameter joint trigger condition is met according to the luminance change rate, the color change rate, the transmission delay deviation rate and the adjusted abnormality judging threshold; the joint trigger condition includes a two-parameter correlation trigger, a transmission delay and physical parameter coupling trigger or a multi-parameter trend consistency trigger;

[0044] a collection state adjusting module configured to increase the collection frequency of the running state parameter and shorten the delay sampling interval for the region meeting the trigger condition, and to decrease the corresponding collection frequency and interval for the region not meeting the condition;

[0045] a data extracting module configured to pre-process the running state parameter, the transmission delay response time and the environmental temperature and humidity parameter collected in real time, and extract the region luminance attenuation trend, the color drift amount, the delay deviation trend, the parameter correlation degree feature and the global fluctuation entropy feature;

[0046] a life prediction module configured to input the feature parameter into a long short-term memory neural network model with an attention mechanism, and output a remaining life prediction value.

[0047] In a third aspect, the present application provides an electronic device, comprising:

[0048] at least one processor; and

[0049] a memory in communication with the at least one processor; wherein

[0050] the memory stores instructions executable by the processor, and the instructions are configured to be executed by the processor to implement a full-motion analog projector life prediction method.

[0051] The present application has the following advantages:

[0052] The present application introduces a regional multi-parameter collection strategy in the dark field mode, effectively capturing early aging signals such as luminance attenuation and color drift that are difficult to detect in the conventional working mode, thereby improving the early fault recognition ability. Through an intelligent trigger mechanism based on the multi-parameter change rate and the dynamic threshold, the system can adaptively adjust the collection frequency of the key region, optimizing the system resource allocation while ensuring data effectiveness. By fusing the transmission delay response time, a key performance indicator, and its deviation from the standard benchmark, and combining the correlation features between parameters and the global stability features, the feature dimension of the model input is enriched, and the theoretical basis for life prediction is enhanced. By introducing a deep learning model with an attention mechanism, focused learning of the degradation features of high-impact areas is realized, and the sensitivity of the model to local aging phenomena is improved. Through a three-level threshold self-modification mechanism, the abnormality judging standard can be dynamically adjusted according to the device aging state, historical fluctuation rules and environmental conditions, enhancing the adaptability and reliability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0053] Other features, objects, and advantages of the application will become more apparent from the following detailed description of non-limiting embodiments thereof, when read in connection with the following accompanying drawings:

[0054] Figure 1 is a flowchart of a full-motion simulator projector life prediction method of the present application. DETAILED DESCRIPTION

[0055] The application will be further described in details below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related application, but not to limit the application. In addition, it should be noted that only the parts related to the application are shown in the drawings for ease of description.

[0056] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and embodiments.

[0057] In the first embodiment of the present application, a full-motion simulator projector life prediction method is provided, which comprises:

[0058] Step S10, starting the dark field mode of the projector, dividing the projection picture into multiple independent regions, and collecting the running state parameters, transmission delay response time and environmental temperature and humidity parameters of each region;

[0059] Step S20, calculating the brightness change rate and color change rate according to the running state parameters of each region, and comparing the current transmission delay response time with the standard reference value to obtain the transmission delay deviation rate;

[0060] Step S30, dynamically adjusting the abnormality judgment threshold according to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability;

[0061] Step S40, judging whether the cross-parameter joint triggering condition is met according to the brightness change rate, the color change rate, the transmission delay deviation rate and the adjusted abnormality judgment threshold; the joint triggering condition includes double-parameter correlation triggering, transmission delay and physical parameter coupling triggering or multi-parameter trend consistency triggering;

[0062] Step S50, increasing the collection frequency of the running state parameters of the region meeting the triggering condition and shortening the delay sampling interval, and reducing the corresponding collection frequency and interval of the region not meeting the condition;

[0063] Step S60, the running state parameters, transmission delay response time and environmental temperature and humidity parameters collected in real time are preprocessed, and the regional brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature are extracted.

[0064] Step S70, the feature parameters are input into the long short-term memory neural network model with attention mechanism, and a remaining life prediction value is output.

[0065] In order to more clearly describe the life prediction method of the full-motion simulator projector of the present application, the following will be combined with Figure 1 The steps in the embodiment of the present application are described in detail as follows:

[0066] Step S10, the dark field mode of the projector is started, the projection picture is divided into multiple independent regions, and the running state parameters, transmission delay response time and environmental temperature and humidity parameters of each region are collected.

[0067] The running state parameters at least include brightness parameters, color parameters, current parameters and temperature parameters.

[0068] In the implementation process of the present application, first, the dark field mode instruction is sent to the projector through the projector control interface (such as RS232 interface), so that the projector closes the active light source, only retains the background dim light or starts the special dark field calibration light source, ensures that the projection picture is in the dark field state, and the background brightness is stabilized at 0.1 cd / m².

[0069] Then, the projection picture is divided into 256 independent regions according to 16*16 grid, and the size of each region is 1 / 256 of the physical size of the projection picture; subsequently, the running state parameters, transmission delay response time and environmental temperature and humidity parameters are collected through the sensors arranged in each region of the projector, wherein one high-sensitivity photoelectric sensor is arranged at the center and four corners of each region for collecting brightness parameters, and the measurement range is 0.001-100 cd / m 2The system has an accuracy of ±1%. RGB three-color recognition sensors are placed at the edge of each area to collect color parameters, with a sampling frequency of 1kHz. Hall effect current sensors are connected in series in the light source driving circuit corresponding to each area to collect current parameters, with a measurement range of 0-10A and an accuracy of ±0.5%. Two thermocouple temperature sensors are attached near the light source module and optical lens in each area to collect temperature parameters, with a measurement range of -40-150℃ and an accuracy of ±0.3℃. Integrated temperature and humidity sensors are installed at key locations such as the exterior of the projector body to collect ambient temperature and humidity parameters, with a temperature measurement range of -40-85℃ and a humidity measurement range of 0-100%RH, and a temperature accuracy of ±0.2℃ and a humidity accuracy of ±2%RH. At the same time, the transmission delay response time is collected through a QTG test module, which is connected to the projector's signal input and display output terminals.

[0070] Finally, the brightness value, color parameters (R, G, B), current value, temperature value, transmission delay response time, and ambient temperature and humidity parameters of each area are collected at an initial frequency of 1 time / minute. The data is received in real time through the software layer, and a moving average algorithm is used for filtering. Outliers are identified using the 3σ criterion and corrected by averaging the five adjacent valid values ​​to complete the preliminary data preprocessing.

[0071] Step S20: Calculate the brightness change rate and color change rate based on the operating status parameters of each area, and compare the current transmission delay response time with the standard reference value to obtain the transmission delay deviation rate;

[0072] In implementation of this invention, at the beginning of each acquisition cycle, for each independent region, the brightness change rate, color change rate, and transmission delay deviation rate are calculated based on the currently acquired operating status parameters; for the calculation of the brightness change rate, the brightness value of region i in the k-th acquisition is taken. Compared with the brightness value collected previously According to the formula Calculations are performed to obtain the percentage value of the brightness change rate of the region in the current period; for the calculation of the color change rate, the RGB color parameter value of region i in the kth acquisition is taken. Corresponding value to the previous collection Calculate the absolute value of the relative change for each channel, i.e.:

[0073] Then take the maximum value among the three and apply the formula. Calculations are performed to obtain the percentage value of the color change rate of the area in the current cycle;

[0074] For the calculation of the transmission delay deviation rate, first, the current transmission delay response time D collected by the QTG test module in real time is obtained, and the standard reference value, i.e., the main QTG value D, preset and stored in the system is called main The value is obtained by averaging multiple QTG tests in the initial state of the projector, and then calculated according to the formula to obtain the transmission delay deviation rate percentage value of the current period.

[0075] All calculation processes are automatically completed by the software layer in the system, and the calculation results are stored in the database together with the corresponding region identification and time stamp for subsequent threshold value judgment and feature extraction.

[0076] Step S30, dynamically adjusting the abnormality judgment threshold according to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability;

[0077] In this embodiment, step S30 includes:

[0078] Step S31, selecting a corresponding first correction coefficient based on different aging stages to which the cumulative working time of the device belongs, wherein the aging stages are divided according to the working time range;

[0079] Step S32, calculating a second correction coefficient based on the difference between the historical parameter fluctuation amplitude and the average fluctuation standard deviation of the device;

[0080] Step S33, determining a third correction coefficient based on whether the environmental temperature and humidity fluctuation data exceeds the stability critical value;

[0081] Step S34, multiplying the initial threshold value by the first correction coefficient, the second correction coefficient and the third correction coefficient to obtain the final abnormality judgment threshold after dynamic adjustment.

[0082] Wherein, the difference between the historical parameter fluctuation amplitude and the average fluctuation standard deviation of the device is subjected to ratio operation with the average fluctuation standard deviation of the device, and the operation result is mapped into a preset coefficient range to obtain the second correction coefficient.

[0083] In the implementation of the present application, the system first reads the cumulative working time t (unit: hour) of the projector from the equipment history database, and judges according to the preset aging stage threshold value. If t≤5000 hours, it is determined that the equipment is in a low aging stage, the first correction coefficient k1 is 1.0; if 5000 hours<t≤10000 hours, it is determined that the equipment is in a middle aging stage, k1 is 0.8; if t>10000 hours, it is determined that the equipment is in a high aging stage, k1 is 0.6; then, the system calculates the standard deviation of the change rate sequence of each parameter (luminance change rate, color change rate, transmission delay deviation rate) in the past 30 days, respectively denoted as , as the historical parameter fluctuation amplitude, and simultaneously reads the corresponding parameter average fluctuation standard deviation of the projector of this type from the system configuration library 、 、 For luminance change rate threshold correction, the second correction coefficient of the luminance change rate is calculated, which is realized by calculating the ratio of the difference between the current fluctuation standard deviation and the average fluctuation standard deviation to the average fluctuation standard deviation, and mapping the result to the interval [0.8, 1.2];

[0084] Similarly, the of the color change rate, and the of the transmission delay deviation rate are calculated; then, the system reads the recent environmental temperature and humidity data, calculates the environmental temperature fluctuation standard deviation σ EnvT and the humidity fluctuation standard deviation σ H , if σ EnvT >2℃ or σ H >5%RH, it is determined that the environment is unstable, and the third correction coefficient k3 is 1.1, otherwise k3 is 1.0;

[0085] Finally, the system obtains the initial set abnormal judgment threshold reference value, i.e. the initial threshold value of luminance change rate RL0 base =3%, the initial threshold value of color change rate RC0 base =2%, the initial threshold value of transmission delay deviation rate RD0 base =5%, and dynamically adjusts through multiplication operation to obtain the final dynamic abnormal judgment threshold:

[0086] The final threshold value of luminance change rate RL0=RL0 base ×k1×k2 RL ×k3, the final threshold value of color change rate RC0=RC0 base ×k1×k2 RC ×k3, and the final threshold value of transmission delay deviation rate RD0=RD0 base ×k1×k2 RDXk3; all the calculated final thresholds will be applied to the subsequent joint trigger condition judgment logic.

[0087] Step S40, according to the luminance change rate, color change rate and transmission delay deviation rate and the adjusted abnormality judgment threshold, it is judged whether the cross-parameter joint trigger condition is met; the joint trigger condition includes double-parameter association trigger, transmission delay and physical parameter coupling trigger or multi-parameter trend consistency trigger;

[0088] In this embodiment, according to the luminance change rate, color change rate and transmission delay deviation rate and the adjusted abnormality judgment threshold, it is judged whether the cross-parameter joint trigger condition is met, and the method is:

[0089] The double-parameter association trigger condition is that the change rate or deviation rate of any two parameters simultaneously reaches the set proportion of its respective abnormality judgment threshold;

[0090] The transmission delay and physical parameter coupling trigger condition is that the transmission delay deviation rate reaches the set proportion of its abnormality judgment threshold, and the current change rate or temperature change rate of the corresponding region exceeds the respective set fluctuation limit value;

[0091] The multi-parameter trend consistency trigger condition is that in the continuous multiple acquisition periods, the luminance change rate, color change rate and transmission delay deviation rate all show a monotone increasing trend, and the change rate of at least two parameters exceeds the set proportion of its abnormality judgment threshold.

[0092] In this embodiment, the system performs the judgment logic of the three joint trigger conditions in parallel, and for each independent region, the system reads the current calculated luminance change rate RL(i, k), color change rate RC(i, k), transmission delay deviation rate RD bias(i，k) and the dynamically adjusted final abnormality judgment threshold RL0, RC0, RD0 in real time;

[0093] For the double-parameter association trigger condition, the system judges whether the following any combination is met at the same time: the luminance change rate RL(i, k) reaches 80% and above of its threshold RL0 and the color change rate RC(i, k) reaches 80% and above of its threshold RC0, that is, RL(i, k)≥0.8×RL0 and RC(i, k)≥0.8×RC0; or

[0094] The luminance change rate RL(i, k) reaches 80% and above of its threshold RL0 and the transmission delay deviation rate RD bias(i，k) reaches 80% and above of its threshold RD0, that is, RL(i, k)≥0.8×RL0 and RD bias(i，k) ≥0.8×RD0; or

[0095] The color change rate RC(i,k) reaches 80% or more of its threshold RC0 and the transmission delay deviation rate RD bias(i，k) Reaching 80% or more of its threshold RD0, i.e., RC(i,k)≥0.8×RC0 and RD bias(i，k) ≥0.8×RD0;

[0096] For the coupling trigger condition between transmission delay and physical parameters, the system first determines the transmission delay deviation rate RD. bias(i，k) Whether it reaches 70% or more of its threshold RD0, i.e., RD bias(i，k) If the value is ≥0.7×RD0, then further determine whether the physical parameters of the region are abnormal, i.e., calculate the rate of change of current in the region. If RI(i,k)≥5%, then an abnormal current fluctuation is determined, or the temperature change rate of the region is calculated. If RT(i,k)≥3%, then abnormal temperature fluctuation is determined. This condition is triggered when the transmission delay deviation condition and any physical parameter abnormal fluctuation condition are met simultaneously.

[0097] For the multi-parameter trend consistency trigger condition, the system retrieves the brightness change rate sequence RL(i,k-4)...RL(i,k), color change rate sequence RC(i,k-4)...RC(i,k), and transmission delay deviation rate sequence RD for the most recent five consecutive acquisition cycles (k-4,k-3,k-2,k-1,k) in the region. bias(i，k-4) ...RD bias(i，k) First, determine whether each of the three sequences exhibits a monotonically increasing trend, meaning that for any two adjacent time points in the sequence, the value at the later time point is greater than or equal to the value at the previous time point. Then, determine whether, within this continuous period, at least two parameters (brightness and color, brightness and delay, or color and delay) have a rate of change exceeding 60% of their respective final anomaly thresholds (RL0, RC0, RD0) at the current time (time k). Specifically, RL(i, k) > 0.6 × RL0 and RC(i, k) > 0.6 × RC0, or RL(i, k) > 0.6 × RL0 and RD0 > RC0. bias(i，k) >0.6×RD0, or RC(i,k)>0.6×RC0 and RD bias(i，k) >0.6×RD0;

[0098] If any one of the above three combined triggering conditions is met, the system determines that the region meets the cross-parameter combined triggering condition and generates a corresponding triggering flag signal to drive the acquisition frequency adjustment operation in step S50.

[0099] Step S50, for the region that meets the trigger condition, increase the collection frequency of its running state parameters and shorten the delay sampling interval, for the region that does not meet the condition, reduce the corresponding collection frequency and interval;

[0100] In this embodiment, the system dynamically adjusts the collection strategy of each region according to the cross-parameter joint trigger condition judgment result generated in step S40; for the region judged to meet any trigger condition, the system marks it as a "high change rate region", and through the control of the hardware layer and the software layer of the data collection module, the collection frequency of the running state parameters (including brightness, color, current, temperature) of the region is increased from the initial 1 time / minute to 5 times / minute, and the sampling interval of the transmission delay response time of the region is shortened from the initial synchronization with the running state parameters (i.e. 1 time / minute) to 1 second, which is realized by sending a control instruction to the QTG test module to make it perform testing and return data at a higher frequency;

[0101] For the region that is not judged to meet any trigger condition, the system marks it as a "low change rate region", and reduces the collection frequency of the running state parameters of the region from the initial 1 time / minute to 1 time / 10 minutes, and extends the sampling interval of the transmission delay response time of the region to 30 seconds; the frequency adjustment process is real-time and automatic, and the system maintains a region state mapping table, continuously updates the latest marked state of each region and its current effective collection frequency and sampling interval settings;

[0102] When a region no longer meets any trigger condition due to changes in subsequent collected data, the system removes its marked state from "high change rate region" and automatically restores its collection frequency and sampling interval to low frequency settings, and vice versa, when a previously untriggered region meets the condition in a new round of judgment, the collection frequency is immediately increased and the sampling interval is shortened; in addition, the system records the adjustment time, region identification, frequency and interval values before and after adjustment, and the specific condition type that triggers the adjustment, forms a log for operation and maintenance analysis and audit.

[0103] Step S60, pre-process the real-time collected running state parameters, transmission delay response time and environmental temperature and humidity parameters, and extract the region brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature;

[0104] Specifically, step S60 includes:

[0105] Step S61, pre-process the real-time collected running state parameters, transmission delay response time and environmental temperature and humidity parameters, the pre-processing includes identifying abnormal values, correcting abnormal values and filtering processing;

[0106] Step S62, based on the pre-processed data, extracting regional feature parameters and global feature parameters, wherein:

[0107] Step S63, the regional feature parameters are obtained by calculating the average attenuation of the brightness value of each region, the average value of the color change rate and the average value of the transmission delay deviation rate;

[0108] Step S64, the parameter correlation degree feature is obtained by calculating the correlation coefficient of the transmission delay deviation rate and the current change rate and the correlation coefficient of the temperature change rate;

[0109] Step S65, the global fluctuation entropy feature is obtained by calculating the sample entropy of the brightness change rate, color change rate and delay deviation rate sequence of all regions.

[0110] In this embodiment, the system first pre-processes the original data collected in real time, and for each parameter sequence (including the brightness, color, current, temperature, transmission delay response time and environment temperature and humidity of each region), adopts 3σ criterion for outlier identification, that is, the mean value μ and the standard deviation σ of the current data window (recent 100 sampling points) of each parameter are calculated, if a data point x satisfies |x-μ|>3σ, it is marked as an outlier;

[0111] For the identified outliers, the average value of the adjacent 5 valid values is used for correction, and the specific formula is , wherein is the nearest valid data point before and after the outlier position k, if the adjacent valid points are insufficient, linear interpolation method is used for estimation;

[0112] After completing the outlier correction, a sliding average filter with a window size of 5 is applied to each parameter sequence for smoothing processing, and the filter output value y k is calculated according to the formula , wherein to is the current point k and the adjacent original data values before and after it;

[0113] After the pre-processing is completed, the system extracts the regional feature parameters based on the filtered data, for each region i, the brightness attenuation trend value is calculated, wherein L(i,0) is the initial brightness reference value of the region, L(i,t) is the brightness value collected for the tth time, and k is the current total collection number; the color drift amount is calculated, wherein RC(i,t) is the color change rate collected for the tth time;

[0114] The delay deviation trend is calculated, wherein is the transmission delay deviation rate of the t-th acquisition; meanwhile, the parameter correlation degree feature, i.e., the Pearson correlation coefficient of the transmission delay deviation rate and the current change rate, is calculated wherein and are the sequences and the mean of RI(i, t), RI(i, t) being the current change rate;

[0115] Similarly, the Pearson correlation coefficient Corr {D-T}(i) of the transmission delay deviation rate and the temperature change rate is calculated; at the global level, the system calculates the global fluctuation entropy feature, first splicing the luminance change rate sequence {RL(1, t)...RL(256, t)}, the color change rate sequence {RC(1, t)...RC(256, t)} and the delay deviation rate sequence {RD bias(1，t) ...RD bias(256，t)} of all regions into three long sequences respectively, and then calculating the entropy value of each long sequence based on the sample entropy algorithm, setting the embedding dimension m = 2 and the tolerance r = 0.2 times the standard deviation of the sequence, finally obtaining three entropy values Entropy RL , Entropy RC , Entropy RD , and taking the average value thereof as the global fluctuation entropy feature Entropy fluct ; all the extracted feature parameters are finally organized into a feature vector and time-stamped into a feature database to provide input for the subsequent life prediction model.

[0116] Step S70, input the feature parameters into the long short-term memory neural network model with attention mechanism, and output the remaining life prediction value.

[0117] The system organizes the feature parameters extracted and preprocessed in step S60 into an input vector of fixed format, the dimension of the vector being 1028, which is specifically spliced from the feature vectors of 256 regions and 4 global feature vectors; the feature vector of each region contains 4 parameters, i.e., the luminance attenuation trend value A L(i) , the color drift amount A C(i) , the delay deviation trend A D(i) and the correlation coefficient Corr {D-I}(i) of the transmission delay deviation rate and the current change rate, so that 256 regions contribute 1024 feature dimensions; the remaining 4 global feature dimensions are respectively: the transmission delay response time change rate RD(k), the average current I avg , the average temperature T avg and the global fluctuation entropy Entropy fluct ;

[0118] Before feeding the input vector into the prediction model, the system calls the pre-trained normalization parameters (including the historical minimum value x min and maximum value x max of each feature dimension) to normalize each feature value of the input vector to the interval [0, 1] using the formula x' = (x - x min ) / (x max - x min ); the normalized input vector is then fed into an improved Long Short-Term Memory (LSTM) model with attention mechanism;

[0119] The network structure contains an input layer (1028 neurons), three hidden layers (each containing 64 LSTM neurons, all using ReLU activation function), and an output layer (1 linearly activated neuron); during the forward propagation process, the network first applies an attention mechanism to the 1024-dimensional features of the 256 regions; specifically, the system calculates the feature importance score S i of each region i through a single-layer neural network, using the formula S i = W s ·[A L(i) ,A C(i) ,A D(i) ,Corr {D-I}(i) ]+b s , where W s and b s are the weight matrix and bias vector obtained through training;

[0120] Subsequently, the importance scores of all regions are normalized using the Softmax function to obtain the attention weight of each region; the weighted regional feature vector is calculated using the formula , and this 4-dimensional vector is then concatenated with the aforementioned 4-dimensional global feature vector to form an 8-dimensional comprehensive feature vector; this comprehensive feature vector is then sequentially passed through the three LSTM hidden layers for sequence modeling and transformation, and finally a normalized remaining life prediction value ŷ is generated by the linear output layer; the system finally performs a denormalization operation on this output value using the formula , where and are the maximum and minimum values of the actual remaining life in the training set, to obtain the final remaining life prediction value with physical meaning (unit: hours); this prediction value is output to the user interface along with its timestamp and confidence interval (estimated by the model during the training phase) and stored in the prediction result database.

[0121] In this embodiment, after step S70, there is also a step S80:

[0122] A maintenance warning is issued when the predicted value is below a preset threshold.

[0123] In this embodiment, the system monitors the remaining life prediction value output by step S70 in real time, and compares the value with a preset maintenance threshold; the maintenance threshold is a fixed value preset according to the device model, historical maintenance records and criticality and stored in the system configuration library, for example, it can be set to 500 hours; when the system detects that the remaining life prediction value is less than or equal to the maintenance threshold, a maintenance warning process is triggered immediately; first, the system generates a structured warning message, which contains the following fields: warning unique identifier (generated by timestamp and device ID hash), triggering time, device number, predicted remaining life value, preset maintenance threshold, and specific condition description of triggering; then, the system sends the warning information through integrated multiple communication interfaces in parallel, the way includes: real-time display and prompt in the form of red pop-up window and sound alarm in the graphical user interface (GUI) of the system, at the same time, calling the API interface of the enterprise internal mail system, sending the warning information to the mailbox of the device administrator and maintenance engineer according to the preset mail template format, and pushing the core information (such as device number and remaining life) of the warning content to the mobile terminal of the relevant responsible personnel in the form of short message through the short message gateway (SMSGateway); in addition, the warning message will be automatically recorded to the maintenance work order database, and an emergency maintenance work order with the status of "to be handled" is generated, which is automatically assigned to the corresponding maintenance team; the system will continue to monitor the warning state, if the maintenance work order state is not updated to "handled" or "confirmed" within a preset time (such as 24 hours), the system will automatically trigger the warning escalation process, that is, repeat the above notification process and additionally copy the higher level management personnel; all warning triggering, notification sending, work order generation and state updating operations are recorded in the system audit log in detail to ensure the traceability of the process.

[0124] Although the above embodiment describes each step in the above order, those skilled in the art can understand that, in order to achieve the effect of the embodiment, the different steps do not have to be executed in such order, they can be executed at the same time (in parallel) or in reverse order, and these simple changes are within the protection scope of the present application.

[0125] The second embodiment of the present application provides a full-motion simulator projector life prediction system for realizing a full-motion simulator projector life prediction method, and the system comprises:

[0126] The data acquisition module is configured to start the dark field mode of the projector, divide the projection picture into multiple independent areas, and collect the running state parameters, transmission delay response time and environmental temperature and humidity parameters of each area;

[0127] a data calculation module configured to calculate a luminance change rate and a color change rate according to the operating state parameters of each region, and to obtain a transmission delay deviation rate by comparing the current transmission delay response time with a standard reference value;

[0128] a threshold updating module configured to dynamically adjust an abnormality determination threshold according to the cumulative working time of the device, the historical parameter fluctuation amplitude, and the environmental stability;

[0129] a trigger condition determination module configured to determine whether a cross-parameter joint trigger condition is met according to the luminance change rate, the color change rate, the transmission delay deviation rate, and the adjusted abnormality determination threshold; the joint trigger condition includes a double-parameter correlation trigger, a transmission delay and physical parameter coupling trigger, or a multi-parameter trend consistency trigger;

[0130] a collection state adjustment module configured to increase the collection frequency of the operating state parameters and shorten the delay sampling interval for the region that meets the trigger condition, and to reduce the corresponding collection frequency and interval for the region that does not meet the condition;

[0131] a data extraction module configured to preprocess the real-time collected operating state parameters, transmission delay response time, and environmental temperature and humidity parameters, and to extract region luminance attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature, and global fluctuation entropy feature;

[0132] a life prediction module configured to input the feature parameters into a long short-term memory neural network model with attention mechanism, and to output a remaining life prediction value.

[0133] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related description of the system described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0134] It should be noted that the full-motion analog projector life prediction system provided in the above embodiments is only exemplified by the division of the above functional modules, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the modules or steps in the embodiments of the present application are further decomposed or combined, for example, the modules of the above embodiments can be combined into one module, or can be further split into multiple sub-modules to complete all or part of the functions described above. The names of the modules and steps involved in the embodiments of the present application are only for distinguishing each module or step, and should not be considered as an improper limitation of the present application.

[0135] The electronic device of the third embodiment of the present application comprises:

[0136] at least one processor; and

[0137] a memory in communication with the at least one processor; wherein

[0138] The memory stores instructions executable by the processor for execution by the processor to implement the full-movement simulator projector life prediction method.

[0139] The fourth embodiment of the application is a computer readable storage medium storing computer instructions for execution by the computer to implement the full-movement simulator projector life prediction method.

[0140] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the storage device and the processing device described above and the related descriptions can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0141] Those skilled in the art should appreciate that the modules, method steps of the examples described in connection with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. The software modules, method steps corresponding programs can be placed in random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. In order to clearly illustrate the interchangeability of electronic hardware and software, the components and steps of the examples have been generally described in the foregoing description. Whether the functions are performed in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0142] The terms "first", "second", and the like are used to distinguish similar objects, not to describe or indicate a particular order or sequence.

[0143] The term "comprising" or any other similar term is intended to encompass non-exclusive inclusion, so that a process, method, article or device / apparatus including a series of elements includes not only those elements, but also other elements not explicitly listed, or inherent to the process, method, article or device / apparatus.

[0144] The technical scheme of the present application has been described in combination with the preferred embodiments shown in the drawings, but it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical schemes after the changes or replacements will all fall within the protection scope of the present application.

Claims

1. A method for predicting the lifetime of a full motion simulator projector, characterized in that, The method comprises: starting a dark field mode of the projector, dividing a projection picture into multiple independent regions, and collecting running state parameters, transmission delay response time and environmental temperature and humidity parameters of each region; calculating a brightness change rate and a color change rate according to the running state parameters of each region, and comparing the current transmission delay response time with a standard reference value to obtain a transmission delay deviation rate; dynamically adjusting an abnormality judgment threshold according to a cumulative working time length of the device, a historical parameter fluctuation amplitude and environmental stability; judging whether a cross-parameter joint triggering condition is met according to the brightness change rate, the color change rate, the transmission delay deviation rate and the adjusted abnormality judgment threshold; the joint triggering condition comprises a double-parameter correlation triggering, a transmission delay and physical parameter coupling triggering or a multi-parameter trend consistency triggering; increasing a collection frequency of the running state parameters and shortening a delay sampling interval for a region meeting the triggering condition, and decreasing a corresponding collection frequency and interval for a region not meeting the condition; preprocessing the real-time collected running state parameters, transmission delay response time and environmental temperature and humidity parameters, and extracting a region brightness attenuation trend, a color drift amount, a delay deviation trend, a parameter correlation degree feature and a global fluctuation entropy feature; inputting the feature parameters into a long short-term memory neural network model with an attention mechanism, and outputting a remaining life prediction value.

2. The method of claim 1, wherein, The method for dynamically adjusting the abnormality judgment threshold according to the cumulative working time length of the device, the historical parameter fluctuation amplitude and the environmental stability is as follows: selecting a corresponding first correction coefficient based on different aging stages to which the cumulative working time length of the device belongs, wherein the aging stages are divided according to a working time length range; calculating a second correction coefficient based on a difference between the historical parameter fluctuation amplitude and a device average fluctuation standard deviation; determining a third correction coefficient based on whether environmental temperature and humidity fluctuation data exceeds a stability critical value; multiplying an initial threshold value by the first correction coefficient, the second correction coefficient and the third correction coefficient to obtain a final abnormality judgment threshold after dynamic adjustment.

3. The method of claim 2, wherein the method further comprises: The difference between the historical parameter fluctuation amplitude and the device average fluctuation standard deviation is subjected to ratio operation with the device average fluctuation standard deviation, and the operation result is mapped into a preset coefficient range to obtain the second correction coefficient.

4. The method of claim 1, wherein, The method for judging whether the cross-parameter joint triggering condition is met according to the brightness change rate, the color change rate, the transmission delay deviation rate and the adjusted abnormality judgment threshold is as follows: the double-parameter correlation triggering condition is that the change rates or deviation rates of any two parameters simultaneously reach a set proportion of their respective abnormality judgment thresholds; the transmission delay and physical parameter coupling triggering condition is that the transmission delay deviation rate reaches a set proportion of its abnormality judgment threshold, and the current change rate or temperature change rate of the corresponding region exceeds a respective set fluctuation limit value; the multi-parameter trend consistency triggering condition is that, in a plurality of continuous collection periods, the brightness change rate, the color change rate and the transmission delay deviation rate all show a monotone increasing trend, and the change rates of at least two parameters exceed a set proportion of their respective abnormality judgment thresholds.

5. The method of claim 1, wherein, The running state parameters, transmission delay response time and environmental temperature and humidity parameters collected in real time are preprocessed, and regional brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature are extracted, and the method is as follows: The running state parameters, transmission delay response time and environmental temperature and humidity parameters collected in real time are preprocessed, and the preprocessing includes identifying abnormal values, correcting abnormal values and filtering processing; Based on the preprocessed data, regional feature parameters and global feature parameters are extracted, wherein: The regional feature parameters are obtained by calculating the average attenuation amount of the brightness values of each region, the average value of the color change rate and the average value of the transmission delay deviation rate; The parameter correlation degree feature is obtained by calculating the correlation coefficient of the transmission delay deviation rate and the current change rate and the correlation coefficient of the temperature change rate; The global fluctuation entropy feature is obtained by calculating the sample entropy of the sequence of all regional brightness change rates, color change rates and delay deviation rates.

6. The method of claim 5, wherein the method further comprises: The feature parameters are input into a long short-term memory neural network model with attention mechanism, and a residual life prediction value is output, and the method is as follows: The regional feature parameters and global feature parameters are combined into a model input vector; The input vector is input into a pre-trained long short-term memory neural network model; The long short-term memory neural network model calculates the weight of each regional feature parameter through the attention mechanism layer inside it, and performs weighted summation on the regional feature parameters to focus on the features of the high-impact region; The processing layer of the long short-term memory neural network model operates on the weighted features to output a prediction value representing the normalized residual life; The prediction value of the normalized residual life is denormalized to obtain the final residual life prediction value.

7. The method of claim 1, wherein the method further comprises: The running state parameters at least include brightness parameters, color parameters, current parameters and temperature parameters.

8. The method of claim 1, wherein, When the prediction value is lower than the preset threshold, a maintenance warning is issued.

9. A system for predicting the lifetime of a moving map projector, for implementing a method for predicting the lifetime of a moving map projector according to any one of claims 1 to 8, characterized in that, The system comprises: A data acquisition module configured to start the dark field mode of the projector, divide the projection picture into multiple independent regions, and collect the running state parameters, transmission delay response time and environmental temperature and humidity parameters of each region; A data calculation module configured to calculate the brightness change rate, color change rate according to the running state parameters of each region, and obtain the transmission delay deviation rate by comparing the current transmission delay response time with the standard reference value; A threshold updating module configured to dynamically adjust the abnormal judgment threshold according to the cumulative working time of the device, the historical parameter fluctuation amplitude and the environmental stability; A trigger condition judgment module configured to determine whether the cross-parameter joint trigger condition is met according to the brightness change rate, color change rate and transmission delay deviation rate and the adjusted abnormal judgment threshold; the joint trigger condition includes double-parameter correlation trigger, transmission delay and physical parameter coupling trigger or multi-parameter trend consistency trigger; A collection state adjustment module configured to increase the collection frequency of the running state parameters and shorten the delay sampling interval for the regions that meet the trigger condition, and reduce the corresponding collection frequency and interval for the regions that do not meet the condition. a data extraction module configured to preprocess the real-time collected operating state parameters, transmission delay response time and environmental temperature and humidity parameters, and extract regional brightness attenuation trend, color drift amount, delay deviation trend, parameter correlation degree feature and global fluctuation entropy feature; a life prediction module configured to input the feature parameters into a long short-term memory neural network model with attention mechanism, and output a residual life prediction value.

10. An electronic device, comprising: comprising: at least one processor; and a memory connected to the at least one processor in communication; wherein, the memory stores instructions executable by the processor, the instructions being executed by the processor to implement the life prediction method of the full-motion analog projector according to any one of claims 1-8.

Images