Method and apparatus for processing non-uniform time series data, electronic device, and storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHONGQING CHANGAN TECH CO LTD
- Filing Date
- 2023-02-20
- Publication Date
- 2026-06-23
Smart Images
Figure CN116307112B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of prediction technology for non-uniform time series data, and in particular to a method, apparatus, electronic device and storage medium for processing non-uniform time series data. Background Technology
[0002] Classification or regression models based on time-series data are already a relatively mature business scenario. Common time-series tasks include text classification, text sentiment analysis, weather forecasting, and stock forecasting.
[0003] However, for data generated by passenger vehicles, most of the signal data is generated by onboard sensors and uploaded to a server database via vehicle-to-everything (V2X) communication. During this data flow, there are several points where data loss can occur. For example, sensor malfunctions can prevent the acquisition of environmental signals; CAN bus transmission failures can cause missed or duplicate data transmissions; and unstable vehicle-side signals can lead to data loss during remote data interaction. In addition to these reasons, the vehicle-side signal acquisition strategy itself is also quite complex. For instance, different signals are acquired using different methods; some signals are acquired using variable acquisition, while others are acquired periodically. The periods of periodically acquired signals may not be consistent, and different signals may flow through different signal channels.
[0004] Therefore, the actual data generated by private cars is at time t, some types of data are at time t+s, and there is another part of the data. Here, s is an indefinite time period, thus constituting non-uniform time-series data generated by the entire process.
[0005] Non-uniform time-series data cannot be directly used as input data for time-series models. There are two reasons for this: First, at any given moment, only partial feature data is available, and the complete feature input cannot be guaranteed, requiring feature value filling. Second, even assuming that all inputs to the model at any given moment have relevant data, there may be long periods of blank data in the automotive data, such as from when the car was turned off the previous night to when it was turned on the next morning. If this data is directly input into the time-series model, the model lacks the ability to distinguish between different data points, potentially leading to a situation where the previous night's data directly influences the results of the following morning. Summary of the Invention
[0006] This application provides a method, apparatus, electronic device, and storage medium for processing non-uniform time-series data, which solves the problem that non-uniform time-series data cannot be directly used as input data for time-series models. It can predict how changes in vehicle interior temperature, vehicle speed, and light intensity will affect the driver's use of air conditioning and air conditioning temperature settings over a period of time, thereby estimating the time when the driver will use the air conditioning and the temperature setting, and recommending the appropriate air conditioning settings for the driver.
[0007] The first aspect of this application provides a method for processing non-uniform time-series data, including the following steps: acquiring non-uniform time-series data of a current vehicle; preprocessing the non-uniform time-series data, and combining the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data groups for prediction; inputting the multiple data groups for prediction into a pre-built prediction model, and predicting the future state of the current vehicle based on the prediction results.
[0008] Based on the aforementioned technical means, it is possible to predict how changes in vehicle interior temperature, vehicle speed, and light intensity over a period of time will affect the driver's use of air conditioning and air conditioning temperature settings. This allows for the estimation of when the driver will use the air conditioning and the desired temperature settings, and the recommendation of appropriate air conditioning settings for the driver.
[0009] Furthermore, the step of combining the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data sets for prediction includes: obtaining the amount of non-uniform time-series data within multiple preset durations; determining the target number of data entries for each preset duration based on the amount of non-uniform time-series data within the multiple preset durations; and generating the multiple data sets for prediction based on the target number of data entries and the preset durations.
[0010] Based on the aforementioned technical means, data groups are generated according to the amount of non-uniform time-series data and the preset duration, and are used to store non-uniform time-series data.
[0011] Furthermore, generating the plurality of data groups for prediction based on the target number of data entries and the preset duration includes: traversing the preprocessed non-uniform time-series data based on the target number of data entries and the preset duration, and sequentially filling the preprocessed non-uniform time-series data into the target data group; if the duration of the data filled into the target data group exceeds the preset duration, then a new target data group is established, and the remaining data of the preprocessed non-uniform time-series data is filled into the new target data group, until all the preprocessed non-uniform time-series data is filled.
[0012] Based on the above technical means, non-uniform time-series data that are filled into the target data group for a period longer than the preset duration are used to establish a new target data group, ensuring that the lengths of multiple data groups are consistent.
[0013] Furthermore, after the preprocessed non-uniform time-series data are all filled, the method further includes: if the data filled in the last target data group does not meet the target number of data entries, then fill the last target empty data group with 0, so that the data filled in the last target data group meets the target number of data entries.
[0014] Using the aforementioned technical methods, the number of data entries in the data set is kept below the set target number to ensure the consistency of the input data for the prediction model.
[0015] Furthermore, after sequentially filling the preprocessed non-uniform time-series data into the target data group, the method further includes: if the duration of the data filled into the target data group reaches the preset duration, then the first data in the target data group is extracted, and it is determined whether the second data in the target data group and the data to be filled meet the preset duration; if the second data in the target data group and the data to be filled meet the preset duration, then the data to be filled is filled into the target data group, and the second data is extracted as a new first data, until the new second data in the target data group and the new data to be filled do not meet the preset duration, and a new target data group is established.
[0016] Based on the above technical means, the time difference between the first data in the data group and the data to be filled is not greater than the preset time, so as to ensure the consistency of the input data of the prediction model.
[0017] Furthermore, the preprocessing of the non-uniform time-series data includes: performing outlier processing, missing value filling, one-hot encoding, and normalization operations on the non-uniform data according to a preset processing strategy, filling strategy, encoding strategy, and normalization strategy.
[0018] Based on the above technical means, missing data in non-uniform time series data are filled in, and one-hot encoding and normalization of other data are performed to ensure the integrity of the data input for the prediction model.
[0019] A second aspect of this application provides a processing apparatus for non-uniform time-series data, comprising: an acquisition module for acquiring non-uniform time-series data of a current vehicle; a processing module for preprocessing the non-uniform time-series data and combining the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data sets for prediction; and a prediction module for inputting the multiple data sets for prediction into a pre-constructed prediction model and predicting the future state of the current vehicle based on the prediction results.
[0020] Furthermore, the processing module is also configured to: acquire the amount of non-uniform time-series data within multiple preset durations; determine the target number of data entries for each preset duration based on the amount of non-uniform time-series data within the multiple preset durations; and generate the multiple data groups for prediction based on the target number of data entries and the preset durations.
[0021] Furthermore, the processing module, which generates the plurality of data groups for prediction based on the target number of data entries and the preset duration, is further configured to: traverse the preprocessed non-uniform time-series data based on the target number of data entries and the preset duration, and sequentially fill the preprocessed non-uniform time-series data into the target data group; if the duration of the data filled into the target data group exceeds the preset duration, then a new target data group is established, and the remaining data of the preprocessed non-uniform time-series data is filled into the new target data group, until all the preprocessed non-uniform time-series data is filled.
[0022] Furthermore, after the preprocessed non-uniform time-series data has been filled, the processing module is further configured to: if the data filled in the last target data group does not meet the target number of data entries, then fill the last target empty data group with 0, so that the data filled in the last target data group meets the target number of data entries.
[0023] Furthermore, after sequentially filling the preprocessed non-uniform time-series data into the target data group, the processing module is further configured to: if the duration of the data filled into the target data group reaches the preset duration, then extract the first data in the target data group and determine whether the second data in the target data group and the data to be filled meet the preset duration; if the second data in the target data group and the data to be filled meet the preset duration, then fill the data to be filled into the target data group and extract the second data as a new first data, until the new second data in the target data group and the new data to be filled do not meet the preset duration, and establish the new target data group.
[0024] Furthermore, the processing module is also used to perform outlier processing, missing value filling, one-hot encoding, and normalization operations on the non-uniform data according to a preset processing strategy, filling strategy, encoding strategy, and normalization strategy.
[0025] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the non-uniform time-series data processing method as described in the above embodiments.
[0026] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the method for processing non-uniform time-series data as described in the above embodiments.
[0027] Therefore, this application acquires non-uniform time-series data of the current vehicle, preprocesses it, and combines the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data sets for prediction. These multiple data sets are then input into a pre-built prediction model, and the future state of the current vehicle is predicted based on the prediction results. This solves the problem that current non-uniform time-series data cannot be directly used as input data for time-series models. It can predict how changes in in-vehicle temperature, vehicle speed, and light intensity over a period of time affect the driver's use of air conditioning and air conditioning temperature settings, thereby estimating the time the driver will use the air conditioning and the appropriate temperature setting, and recommending the necessary air conditioning settings to the driver.
[0028] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description
[0029] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:
[0030] Figure 1 This is a flowchart of a method for processing non-uniform time-series data according to an embodiment of this application;
[0031] Figure 2 This is a schematic diagram of a time series prediction process based on non-uniform time series data according to an embodiment of this application;
[0032] Figure 3 This is a schematic diagram of the input data reconstruction process of an RNN (Recurrent Neural Network) according to an embodiment of this application;
[0033] Figure 4 This is a schematic diagram of an LSTM (Long Short Term) network structure according to an embodiment of this application;
[0034] Figure 5 This is a schematic diagram illustrating the key reconstruction process in data reconstruction according to an embodiment of this application;
[0035] Figure 6 This is a block diagram of a non-uniform time-series data processing apparatus according to an embodiment of this application;
[0036] Figure 7 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application.
[0037] Explanation of reference numerals in the attached figures: 10-processing device for non-uniform time series data, 100-acquisition module, 200-processing module, 300-prediction module, 701-memory, 702-processor, 703-communication interface. Detailed Implementation
[0038] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.
[0039] The following description, with reference to the accompanying drawings, outlines a method, apparatus, electronic device, and storage medium for processing non-uniform time-series data according to embodiments of this application. Addressing the problem mentioned in the background art that current non-uniform time-series data cannot be directly used as input data for time-series models, this application provides a method for processing non-uniform time-series data. In this method, non-uniform time-series data of the current vehicle is acquired and preprocessed. Following a preset combination strategy, the preprocessed non-uniform time-series data is combined to obtain multiple data sets for prediction. These multiple data sets are then input into a pre-constructed prediction model, and the future state of the current vehicle is predicted based on the prediction results. This solves the problem that current non-uniform time-series data cannot be directly used as input data for time-series models. It can predict how changes in in-vehicle temperature, vehicle speed, and light intensity over a period of time affect the driver's use of the air conditioner and the air conditioner temperature setting, thereby estimating the time the driver should use the air conditioner and the desired temperature setting, and recommending the appropriate air conditioner settings to the driver.
[0040] Specifically, Figure 1 This is a flowchart illustrating a method for processing non-uniform time-series data provided in an embodiment of this application.
[0041] like Figure 1 As shown, the processing method for this non-uniform time-series data includes the following steps:
[0042] In step S101, non-uniform time-series data of the current vehicle is obtained.
[0043] Specifically, such as Figure 2 As shown, embodiments of this application can obtain non-uniform time-series data of the current vehicle generated by the vehicle-mounted sensors from the server database.
[0044] In step S102, the non-uniform time series data is preprocessed, and the preprocessed non-uniform time series data is combined according to a preset combination strategy to obtain multiple data groups for prediction.
[0045] In some embodiments, preprocessing of non-uniform time-series data includes: performing outlier processing, missing value filling, one-hot encoding, and normalization operations on the non-uniform data according to preset processing strategies, filling strategies, encoding strategies, and normalization strategies.
[0046] Because vehicle data is generated at different times, meaning the data required by the model at any given moment is incomplete, relevant data preprocessing is needed for the non-uniform temporal data collected and transmitted. This includes outlier handling, missing value imputation, one-hot encoding, and normalization. For missing non-uniform temporal data, a backward imputation method is used to fill in values, primarily to match the data acquisition method. For non-uniform temporal data without missing values, a pre-defined combination strategy is used for data combination.
[0047] Furthermore, in some embodiments, the preprocessed non-uniform time-series data is combined according to a preset combination strategy to obtain multiple data sets for prediction, including: obtaining the amount of non-uniform time-series data within multiple preset durations; determining the target number of data entries for each preset duration based on the amount of non-uniform time-series data within multiple preset durations; and generating multiple data sets for prediction based on the target number of data entries and the preset duration.
[0048] The preset duration can be a threshold set by the user, a threshold obtained through a limited number of experiments, or a threshold obtained through a limited number of computer simulations. Preferably, the preset duration is set to 5 minutes.
[0049] Specifically, such as Figure 3 As shown, according to the expert system's opinion, changes in the environment within the vehicle for a preset period of time will affect the driver's air conditioning settings.
[0050] The distribution of non-uniform time-series data is statistically analyzed to determine the distribution of data volume within a 5-minute timeframe. The statistical results of the data volume within this 5-minute timeframe are obtained, and the 80% threshold number of data points (i.e., the target number of data points) is identified. This determines how many data points in each group will maintain a time difference of approximately 5 minutes for the entire data set. This number is then used as the input to the RNN / LSTM model. The target number of data points for each group is set to 400. The RNN / LSTM model is as follows: Figure 4 As shown.
[0051] Furthermore, in some embodiments, multiple data groups for prediction are generated based on the target number of data entries and a preset duration, including: based on the target number of data entries and the preset duration, traversing the preprocessed non-uniform time-series data, and sequentially filling the preprocessed non-uniform time-series data into the target data group; if the duration of the data filled into the target data group exceeds the preset duration, a new target data group is established, and the remaining data of the preprocessed non-uniform time-series data is filled into the new target data group, until all the preprocessed non-uniform time-series data is filled.
[0052] Specifically, such as Figure 5 As shown, first, an empty data list is set up, and the target number of data entries and preset duration of each data group are obtained. By traversing the non-uniform time series data that has already been filled with values, an empty target data group is set up. The preprocessed non-uniform time series data is then filled into the empty list in turn. If the time between the added data and the first data entry exceeds 5 minutes, a new target data group needs to be created. The remaining data of the preprocessed non-uniform time series data is then filled into the new target data group. If the target number of data entries in the new target data group is less than 400, zeros need to be padded to reach the new target number of 400 to ensure that the length within the data group remains consistent.
[0053] Furthermore, in some embodiments, after sequentially filling the preprocessed non-uniform time-series data into the target data group, the method further includes: if the duration of the data filled into the target data group reaches a preset duration, then the first data in the target data group is extracted, and it is determined whether the second data in the target data group and the data to be filled meet the preset duration; if the second data in the target data group and the data to be filled meet the preset duration, then the data to be filled is filled into the target data group, and the second data is extracted as a new first data, until the new second data in the target data group and the new data to be filled do not meet the preset duration, and a new target data group is established.
[0054] It should be understood that if the data filling time in the target data group reaches the preset time, the first data in the target data group is taken out, and it is determined whether the second data in the target data group and the data to be filled meet the preset time. If they meet the preset time, the data to be filled is filled into the target data group to form a new data group. If they do not meet the preset time, the second data in the target data group is taken out, and it is determined whether the third data and the data to be filled meet the preset time. This process continues until the data in the target data group and the new data to be filled do not meet the preset time, at which point a new target data group is created.
[0055] Furthermore, in some embodiments, after the preprocessed non-uniform time-series data are all filled, the method further includes: if the data filled in the last target data group does not meet the target number of data entries, then fill the last target empty data group with 0, so that the data filled in the last target data group meets the target number of data entries.
[0056] Specifically, after traversing all preprocessed non-uniform time series data, all data within the target data group are obtained. It is determined that the time from the first data to the last data in the target data group does not exceed 5 minutes. The data volume of each target data group is 400 data. If the data filled in the target data group does not meet the target number of data, 0 is filled into the last empty data in each target data group. The constructed data group is then flattened (required for RNN / LSTM input).
[0057] In step S103, multiple sets of data for prediction are input into a pre-built prediction model, and the future state of the current vehicle is predicted based on the prediction results.
[0058] It should be understood that the input to the pre-built prediction model is a set of predicted data, and the output is the future state of the current vehicle. In this embodiment, multiple sets of data for prediction are input into the pre-built prediction model for training, and the trained prediction model is saved. The model's performance is evaluated on a test set to obtain prediction results, and the future state of the current vehicle is predicted. For example,
[0059] Through the above steps, the model focuses more on data within a specific timeframe set by the expert system, enabling it to make reasonable and accurate predictions about future data based on this data. If the traditional method is used, directly inputting data into a time-series model without considering the time range and length limitations of the data set, the model's focus on a specific timeframe becomes unrestricted. Data from the previous night, or even earlier, might be used to predict the next morning's results, leading to unreliable or erroneous predictions that fail to achieve the intended purpose of automatic air conditioning settings.
[0060] The non-uniform time-series data processing method proposed in this application involves acquiring and preprocessing the non-uniform time-series data of the current vehicle. Following a preset combination strategy, the preprocessed non-uniform time-series data is combined to obtain multiple data sets for prediction. These multiple data sets are then input into a pre-built prediction model, and the future state of the vehicle is predicted based on the prediction results. This solves the problem that current non-uniform time-series data cannot be directly used as input data for time-series models. It can predict how changes in in-vehicle temperature, vehicle speed, and light intensity over a period of time affect the driver's use of the air conditioner and the air conditioner temperature setting, thereby estimating the time the driver should use the air conditioner and the desired temperature setting, and recommending the appropriate air conditioner settings to the driver.
[0061] Next, with reference to the accompanying drawings, a processing apparatus for non-uniform time-series data according to an embodiment of this application is described.
[0062] Figure 6 This is a block diagram of a non-uniform time-series data processing apparatus according to an embodiment of this application.
[0063] like Figure 6 As shown, the non-uniform time series data processing device 10 includes: an acquisition module 100, a processing module 200, and a prediction module 300.
[0064] The acquisition module 100 is used to acquire non-uniform time-series data of the current vehicle; the processing module 200 is used to preprocess the non-uniform time-series data and combine the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data sets for prediction; the prediction module 300 is used to input the multiple data sets for prediction into a pre-built prediction model and predict the future state of the current vehicle based on the prediction results.
[0065] Furthermore, in some embodiments, the processing module 200 is also configured to: acquire the amount of non-uniform time-series data within multiple preset durations; determine the target number of data entries for each preset duration based on the amount of non-uniform time-series data within multiple preset durations; and generate multiple data groups for prediction based on the target number of data entries and the preset durations.
[0066] Furthermore, in some embodiments, multiple data groups for prediction are generated based on the target number of data entries and a preset duration. The processing module 200 is further configured to: based on the target number of data entries and the preset duration, traverse the preprocessed non-uniform time-series data and sequentially fill the preprocessed non-uniform time-series data into the target data group; if the duration of the data filled into the target data group exceeds the preset duration, a new target data group is established, and the remaining data of the preprocessed non-uniform time-series data is filled into the new target data group, until all the preprocessed non-uniform time-series data is filled.
[0067] Furthermore, in some embodiments, after sequentially filling the preprocessed non-uniform time-series data into the target data group, the processing module 200 is further configured to: if the duration of the data filled into the target data group reaches a preset duration, then extract the first data in the target data group and determine whether the second data in the target data group and the data to be filled meet the preset duration; if the second data in the target data group and the data to be filled meet the preset duration, then fill the data to be filled into the target data group and extract the second data as the new first data, until the new second data in the target data group and the new data to be filled do not meet the preset duration, and establish a new target data group.
[0068] Furthermore, in some embodiments, after the preprocessed non-uniform time-series data are all filled, the processing module 200 is further configured to: if the data filled in the last target data group does not meet the target number of data entries, then fill the last target empty data group with 0, so that the data filled in the last target data group meets the target number of data entries.
[0069] Furthermore, in some embodiments, the processing module 200 is also used to: perform outlier processing, missing value filling, one-hot encoding, and normalization operations on non-uniform data according to a preset processing strategy, filling strategy, encoding strategy, and normalization strategy.
[0070] It should be noted that the foregoing explanation of the method embodiment for processing non-uniform time-series data also applies to the non-uniform time-series data processing apparatus of this embodiment, and will not be repeated here.
[0071] The non-uniform time-series data processing apparatus proposed in this application acquires the non-uniform time-series data of the current vehicle, preprocesses it, and combines the preprocessed non-uniform time-series data according to a preset combination strategy to obtain multiple data sets for prediction. These multiple data sets are then input into a pre-built prediction model, and the future state of the current vehicle is predicted based on the prediction results. This solves the problem that current non-uniform time-series data cannot be directly used as input data for time-series models. It can predict how changes in in-vehicle temperature, vehicle speed, and light intensity over a period of time affect the driver's use of the air conditioner and the air conditioner temperature setting, thereby estimating the time the driver should use the air conditioner and the desired temperature setting, and recommending the appropriate air conditioner settings to the driver.
[0072] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:
[0073] The memory 701, the processor 702, and the computer program stored on the memory 701 and executable on the processor 702.
[0074] When the processor 702 executes the program, it implements the non-uniform timing data processing method provided in the above embodiments.
[0075] Furthermore, electronic devices also include:
[0076] Communication interface 703 is used for communication between memory 701 and processor 702.
[0077] The memory 701 is used to store computer programs that can run on the processor 702.
[0078] The memory 701 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.
[0079] If the memory 701, processor 702, and communication interface 703 are implemented independently, then the communication interface 703, memory 701, and processor 702 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0080] Optionally, in a specific implementation, if the memory 701, processor 702, and communication interface 703 are integrated on a single chip, then the memory 701, processor 702, and communication interface 703 can communicate with each other through an internal interface.
[0081] The processor 702 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of this application.
[0082] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described method for processing non-uniform time-series data.
[0083] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0084] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0085] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.
[0086] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.
[0087] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
[0088] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.
Claims
1. A method for processing non-uniform time-series data, characterized in that, Includes the following steps: Obtain the non-uniform time-series data of the current vehicle; The non-uniform time series data is preprocessed, and the preprocessed non-uniform time series data is combined according to a preset combination strategy to obtain multiple data groups for prediction. as well as The multiple sets of data used for prediction are input into a pre-built prediction model, and the future state of the current vehicle is predicted based on the prediction results. The preprocessed non-uniform time-series data is combined according to a preset combination strategy to obtain multiple data sets for prediction, including: Acquire the amount of non-uniform time-series data within multiple preset durations; Based on the amount of non-uniform time-series data within the multiple preset time periods, determine the target number of data entries for each preset time period; The multiple data sets for prediction are generated based on the target number of data entries and the preset duration. The generation of the multiple data sets for prediction based on the target number of data entries and a preset duration includes: Based on the target number of data entries and the preset duration, the preprocessed non-uniform time series data is traversed, and the preprocessed non-uniform time series data is sequentially filled into the target data group; If the duration of the data filled into the target data group exceeds the preset duration, a new target data group is created, and the remaining data of the preprocessed non-uniform time series data is filled into the new target data group until all the preprocessed non-uniform time series data is filled.
2. The method according to claim 1, characterized in that, After the preprocessed non-uniform time-series data has been filled, the process further includes: If the data filled in the last target data group does not meet the target number of data entries, then fill the last target empty data group with 0s so that the data filled in the last target data group meets the target number of data entries.
3. The method according to claim 1, characterized in that, After sequentially filling the target data group with the preprocessed non-uniform time-series data, the method further includes: If the data duration for filling the target data group reaches the preset duration, then the first data in the target data group is extracted, and it is determined whether the second data in the target data group and the data to be filled meet the preset duration. If the second data in the target data group and the data to be filled meet the preset duration, then the data to be filled is added to the target data group, and the second data is taken out as the new first data, until the new second data in the target data group and the new data to be filled do not meet the preset duration, and then the new target data group is established.
4. The method according to claim 1, characterized in that, The preprocessing of the non-uniform time-series data includes: According to the preset processing strategy, filling strategy, encoding strategy and normalization strategy, the non-uniform time series data is subjected to outlier processing, missing value filling, one-hot encoding and normalization operations.
5. A processing apparatus for non-uniform time-series data, characterized in that, The processing apparatus is used to implement the method for processing non-uniform time-series data as described in any one of claims 1-4, and the processing apparatus includes: The acquisition module is used to acquire the non-uniform time-series data of the current vehicle; The processing module is used to preprocess the non-uniform time-series data and, according to a preset combination strategy, combine the preprocessed non-uniform time-series data to obtain multiple data sets for prediction; and The prediction module is used to input the multiple sets of data for prediction into a pre-built prediction model, and to predict the future state of the current vehicle based on the prediction results.
6. The apparatus according to claim 5, characterized in that, The processing module is further configured to: Acquire the amount of non-uniform time-series data within multiple preset durations; Based on the amount of non-uniform time-series data within the multiple preset time periods, determine the target number of data entries for each preset time period; The multiple data sets for prediction are generated based on the target number of data entries and the preset duration.
7. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the method for processing non-uniform time-series data as described in any one of claims 1-4.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the method for processing non-uniform time-series data as described in any one of claims 1-4.