Method, device and electronic equipment for weather forecast based on video ringtone

Abstract

The present disclosure provides a weather forecast method and device based on video ringtone and electronic equipment, relating to the technical field of artificial intelligence, the method comprising: performing data fusion processing on weather observation data to obtain standard weather observation data; calculating the weight of each weather index in the standard weather observation data based on historical weather data, optimizing the standard weather observation data according to the weight of each weather index to obtain optimized weather observation data; using sliding window technology to perform time series similarity matching on the current weather observation data to determine the historical weather data segment matched with the current weather observation data; generating weather forecast information according to the subsequent weather data corresponding to the historical weather data segment; generating a video ringtone according to the weather forecast information, and issuing the video ringtone to a target terminal. Through multi-source data fusion and sliding window matching, the prediction accuracy is improved, and combined with the active push of the video ringtone, the weather forecast service with precision, timeliness and wide coverage is realized.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and in particular to a weather forecasting method, apparatus, and electronic device based on video ringback tones. Background Technology

[0002] Currently, weather forecasts are mainly obtained through traditional media (television, radio, etc.), official meteorological channels (websites, mobile apps, etc.), SMS warning services, and mobile internet applications. However, each of these methods has its limitations: traditional media and SMS warnings lack the flexibility and user attention of proactive push notifications; official channels and mobile applications require users to actively search or install specific applications, which poses a barrier to use for older people or those with lower internet literacy; furthermore, weather data from different platforms is independent and lacks cross-platform data integration, making it difficult to provide accurate, unified, and proactive weather forecast push services. Summary of the Invention

[0003] This disclosure provides a weather forecasting method, apparatus, and electronic device based on video ringback tones.

[0004] Firstly, this disclosure provides a weather forecasting method based on video ringback tones, including: Acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data; The weights of each weather indicator in the standard weather observation data are calculated based on historical weather data, and the standard weather observation data are optimized according to the weights of each weather indicator to obtain optimized weather observation data. Acquire current weather observation data, and based on the optimized weather observation data and the historical weather data, use sliding window technology to perform time series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data. Weather forecast information is generated based on subsequent weather data corresponding to the historical weather data segments. A video ringback tone is generated based on the weather forecast information, and the video ringback tone is sent to the target terminal.

[0005] In one possible implementation, the step of calculating the weights of each weather indicator in the standard weather observation data based on historical weather data, and optimizing the standard weather observation data according to the weights of each weather indicator to obtain optimized weather observation data includes: Calculate the error between the observed values ​​of each weather indicator in the standard weather observation data and the corresponding weather indicators in the historical weather data, and calculate the mean square error based on the error; The weights of each weather indicator in the standard weather observation data are calculated based on the mean square error. The standard weather observation data is weighted and optimized according to the weights of each weather indicator to obtain optimized weather observation data.

[0006] In one possible implementation, the step of using the sliding window technique to perform time-series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data includes: Construct a time series matrix; The sliding window is moved across the time series matrix at predetermined steps; After each swipe, calculate the similarity matrix between the current weather observation data and historical weather data within the current window; Record the window position information and similarity matrix obtained from each slide, and generate a similarity set; The similarity set is filtered based on a preset threshold to obtain a candidate similar segment set; The size of the sliding window is increased and the calculation is iterated until the number of elements in the candidate similar segment set is less than or equal to a preset threshold. The historical weather data segment corresponding to the element in the candidate similar segment set is determined as the historical weather data segment that matches the current weather observation data.

[0007] In one possible implementation, generating the video ringback tone based on the weather forecast information includes: Based on the weather type in the weather forecast information, select the corresponding video template from the preset video template library; Dynamically generate text stream content based on the weather forecast information; The text stream content is combined with the video template in real time to generate a video ringback tone; In response to a user's customized operation input through the video ringback tone interface, the text stream content is updated, and user-customized content is generated; the user-customized content includes at least one of customized weather type, push interval, and alarm threshold.

[0008] In one possible implementation, it also includes: When the weather forecast information meets the preset weather conditions, the video ringback tone is actively broadcast to the target terminals in the affected area, and the broadcast is repeated until a confirmation receipt is received.

[0009] In one possible implementation, it also includes: Obtain the family number information set by the user, and push the weather forecast information to the target terminal corresponding to the family number.

[0010] In one possible implementation, the step of acquiring weather observation data from multiple meteorological data sources and performing data fusion processing on the weather observation data to obtain standard weather observation data includes: Weather observation data shall be obtained from at least two of the following sources: meteorological bureau database, meteorological satellite data, meteorological radar data, meteorological observation station data, environmental monitoring station data, early warning information release center, geographic location data platform, and civilian meteorological observation data. The weather observation data is processed into a multidimensional data matrix containing time and location identifiers; the row vectors or column vectors of the multidimensional data matrix represent at least one weather indicator among temperature, humidity, wind speed, wind direction, and air pressure at different time points.

[0011] Secondly, this disclosure provides a weather forecast device based on video ringback tones, comprising: The acquisition module is used to acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data. The optimization module is used to calculate the weight of each weather indicator in the standard weather observation data based on historical weather data, and optimize the standard weather observation data according to the weight of each weather indicator to obtain optimized weather observation data. The matching module is used to acquire current weather observation data, and based on the optimized weather observation data and the historical weather data, to perform time series similarity matching on the current weather observation data using sliding window technology to determine the historical weather data segment that matches the current weather observation data. The generation module is used to generate weather forecast information based on the subsequent weather data corresponding to the historical weather data segment; The delivery module is used to generate a video ringback tone based on the weather forecast information and deliver the video ringback tone to the target terminal.

[0012] Thirdly, this disclosure provides an electronic device, including: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0013] Fourthly, this disclosure provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to perform the method described in the first aspect above.

[0014] The technical solution disclosed in this paper brings at least the following beneficial effects: In the embodiments of this disclosure, weather observation data from multiple meteorological data sources are acquired, and the weather observation data is fused to obtain standard weather observation data. The weights of each weather indicator in the standard weather observation data are calculated based on historical weather data, and the standard weather observation data is optimized according to the weights of each weather indicator to obtain optimized weather observation data. Current weather observation data is acquired, and based on the optimized weather observation data and the historical weather data, a sliding window technique is used to perform time series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data. Weather forecast information is generated based on the subsequent weather data corresponding to the historical weather data segments. A video ringback tone is generated based on the weather forecast information and sent to the target terminal. This enables cross-platform fusion and accurate forecasting of multi-source meteorological data, solving the problem of insufficient forecast accuracy caused by the independence of weather platforms and lack of data fusion in existing technologies. At the same time, by using video ringback tones as the carrier for pushing weather forecast information, weather forecast information is delivered directly to the user's terminal without the need for the user to install any application or actively query it. This solves the problems of access barriers and insufficient user coverage caused by the need for users to actively query or install specific applications in existing technologies.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure, and are not intended to unduly limit this disclosure.

[0017] Figure 1 A flowchart illustrating a first specific embodiment of a weather forecasting method based on video ringback tones provided in this disclosure; Figure 2 A flowchart illustrating a second specific embodiment of a weather forecasting method based on video ringback tones provided in this disclosure; Figure 3 A flowchart illustrating a weather forecasting method based on video ringback tones provided in this embodiment of the present disclosure; Figure 4 This is a schematic diagram of a weather data source provided in an embodiment of the present disclosure; Figure 5 A flowchart for matching historical data is provided in this embodiment of the disclosure; Figure 6This is a schematic diagram of the first jump of a sliding window provided in an embodiment of the present disclosure; Figure 7 A schematic diagram of the second jump of a sliding window provided in an embodiment of this disclosure. Figure 8 A schematic diagram of an S-set with window k=3 provided in an embodiment of this disclosure; Figure 9 This is a schematic diagram of a video ringback tone template provided in an embodiment of the present disclosure; Figure 10 This is a schematic diagram of a weather forecast device based on video ringback tones, provided as an embodiment of the present disclosure. Detailed Implementation

[0018] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0019] The following description, with reference to the accompanying drawings, describes an embodiment of a weather forecasting method, apparatus, and electronic device based on video ringback tones.

[0020] Figure 1 This is a flowchart illustrating a weather forecasting method based on video ringback tones provided in an embodiment of this disclosure. Figure 1 As shown, the method includes the following steps: S101: Acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data.

[0021] In the embodiments of this disclosure, weather observation data from multiple meteorological data sources are acquired. In practical applications, the type and number of meteorological data sources directly affect the accuracy of weather forecasts. The multiple meteorological data sources in this embodiment can be various sources capable of providing weather-related information, such as databases of national meteorological bureaus, remote sensing data collected by meteorological satellites, echo data obtained from meteorological radar scans, measured data recorded by various surface meteorological observation stations, air quality and meteorological correlation data released by the National Environmental Monitoring Center, various meteorological warnings released by the National Early Warning Information Release Center, geographic and meteorological correlation data provided by global geographic location data platforms, and observation data from provincial, municipal, and county-level local meteorological bureaus, and even observation data collected by individual weather enthusiasts using their own equipment. These data sources are distributed across different geographical locations, employ different data acquisition frequencies and data formats, and exhibit heterogeneity.

[0022] After acquiring raw weather observation data from the aforementioned multiple data sources, data fusion processing is required. This process includes several sub-steps: First, the received raw data is cleaned to remove obviously erroneous data. Second, data from different data sources is aligned temporally and spatially, unifying data with inconsistent collection times onto the same timeline and data from different geographic coordinates onto the same geographic grid. Third, data in different formats is converted to a unified data structure. Finally, the cleaned, aligned, and converted data is integrated into a standard weather observation dataset. Standard weather observation data has unified time, location, and weather index identifiers, facilitating subsequent weight calculations and model training.

[0023] Through data fusion processing, this embodiment can fully utilize various available meteorological data resources, overcoming the shortcomings of insufficient data volume and limited coverage of a single data source, and laying a data foundation for subsequent accurate weather forecasting. Standard weather observation data contains complete information in terms of time, space, and weather indicators. It can be understood as a multidimensional data table covering multiple time points and multiple geographical locations, where each cell records the observed value of a specific weather indicator at a specific time and location.

[0024] S102, calculate the weight of each weather indicator in the standard weather observation data based on historical weather data, and optimize the standard weather observation data according to the weight of each weather indicator to obtain optimized weather observation data.

[0025] In the embodiments of this disclosure, the weights of each weather indicator in the standard weather observation data are calculated based on historical weather data, and the standard weather observation data is optimized according to the calculated weights to obtain optimized weather observation data. In this embodiment, historical weather data refers to historical weather records released by a designated authoritative meteorological agency (such as the China Meteorological Administration) that have undergone quality control and correction processing. These historical weather data are considered to be benchmark data with high accuracy and strong reliability. Each weather indicator in the standard weather observation data is compared with the corresponding weather indicator at the corresponding time and location in the historical weather data. Specifically, for the weather indicator of temperature, the difference between the temperature value in the standard weather observation data and the temperature benchmark value in the historical weather data is calculated; for other weather indicators such as humidity, wind speed, wind direction, and air pressure, the difference is also calculated for each indicator. Based on these differences, error statistics for each weather indicator are calculated, such as mean square error. The smaller the mean square error, the closer the value of the weather indicator in the standard weather observation data is to the historical benchmark value, and the higher the reliability; the larger the mean square error, the greater the deviation of the value of the weather indicator from the historical benchmark value, and the lower the reliability. The weight of each weather indicator is calculated based on the mean squared error: the smaller the mean squared error, the greater the weight; the larger the mean squared error, the smaller the weight. In this way, the weight reflects the reliability of different weather indicators in standard weather observation data.

[0026] After obtaining the weights of each weather indicator, the standard weather observation data is weighted and optimized. The core idea of ​​weighted optimization is to give higher importance to weather indicators with high reliability (i.e., indicators with larger weights) in subsequent data use, and to appropriately reduce the influence of weather indicators with low reliability (i.e., indicators with smaller weights). Through weighted optimization, an optimized weather observation data is obtained. This optimized weather observation data has higher accuracy and reliability than the original standard weather observation data because the reliability of each weather indicator has been quantified and adjusted. Weighted optimization can be implemented using various mathematical methods, such as multiplying the values ​​of each weather indicator in the standard weather observation data by the corresponding weight coefficient, or performing a weighted average of observations from different data sources based on the weight coefficients.

[0027] This embodiment enables refined quality assessment and optimization of multi-source weather observation data, effectively improving the accuracy of weather data and providing more reliable input data for subsequent weather forecasts.

[0028] S103: Obtain current weather observation data; based on optimized weather observation data and historical weather data, use sliding window technology to perform time series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data.

[0029] In the embodiments of this disclosure, current weather observation data is obtained, and based on the optimized weather observation data and historical weather data, the time series similarity matching of the current weather observation data is performed using the sliding window technique to determine the historical weather data segment that matches the current weather observation data. Weather forecast information is then generated based on the subsequent weather data corresponding to the historical weather data segment.

[0030] In this embodiment, current weather observation data refers to the latest weather data collected in real time, representing the current weather conditions. Historical weather data refers to a stored and optimized sequence of historical weather observation data, containing weather data at various points in time over a past period. The sliding window technique is a commonly used time series analysis method. Its core idea is to set a fixed-size window on a long sequence, extract subsequences of different time periods by sliding the window, and compare these subsequences with the target sequence. Specifically, in this step, the current weather observation data is first constructed as a time series segment, containing weather data from multiple recent time points. Then, a sliding window of a certain size is defined on the historical weather data sequence, starting from the earliest time point of the historical data sequence and sliding forward step by step according to a predetermined step size. At each new position, a historical weather data subsequence within the window is extracted, and the similarity between this subsequence and the current weather observation data sequence is calculated. The similarity calculation can employ various mathematical methods, such as Euclidean distance, cosine similarity, and Manhattan distance, to quantify the closeness between the two sequences. A smaller similarity value indicates that the two sequences are closer; a larger similarity value indicates that the two sequences are more different. The window position and corresponding similarity value obtained at each sliding step are recorded to generate a similarity set. Then, the similarity set is filtered based on a preset threshold, retaining window positions with similarities less than or equal to the preset threshold to obtain a candidate similar segment set. To improve matching accuracy, an iterative optimization approach can be used: increasing the size of the sliding window, recalculating and filtering similarities within the candidate similar segment set, continuously narrowing the candidate range until the number of elements in the candidate similar segment set is less than or equal to a preset threshold. At this point, the historical weather data segments retained in the candidate similar segment set are considered to be the best match for the current weather observation data. The subsequent weather data immediately following the matched historical weather data segment is extracted as a basis for predicting future weather. For example, if the temperature following the matched historical weather data segment shows an upward trend, it can be predicted that the future weather will be warmer.

[0031] This embodiment can accurately match the similarity of current weather and predict trends based on historical weather patterns, making full use of the meteorological evolution patterns contained in historical data to improve the scientificity and accuracy of weather forecasts.

[0032] S104 generates weather forecast information based on subsequent weather data corresponding to historical weather data segments.

[0033] In the embodiments of this disclosure, video ringback tones are dynamically generated based on weather forecast information. A video ringback tone is a type of video media content played to the calling user during a communication call. Its characteristic is that it does not require the user to install any application or actively query the information; the information is proactively pushed to the user. In this embodiment, the video ringback tone is used as a carrier and pusher of weather forecast information. Based on the weather type in the weather forecast information, a corresponding video template is selected from a preset video template library. The video template library is a pre-created resource collection containing video materials created for different weather types. For example, sunny weather can correspond to a video template with a bright sun and blue sky; rainy weather can correspond to a video template with a visual style of dripping rain and dark clouds; snowy weather can correspond to a video template with a visual style of falling snowflakes and a silver-white landscape; foggy weather can correspond to a video template with a visual style of dense fog and low visibility; and typhoon or windy weather can correspond to a video template with a visual style of swaying trees and howling wind. Each video template can be paired with corresponding background sound effects, such as rain, wind, thunder, and birdsong, to enhance the user experience. Simultaneously, text stream content is dynamically generated based on weather forecast information. This text stream content refers to weather forecast information presented in text form, such as "Today's temperature is 25°C, sunny turning cloudy" or "Heavy rain is expected tomorrow, please take precautions." The text stream content can be updated in real time according to changes in the weather forecast information; for example, as the forecast time approaches or the weather changes, the text content can be dynamically adjusted. The generated text stream content is then combined with the selected video template in real time to generate the final video ringback tone. Real-time combination means that video and text content can be dynamically combined during playback, with text content superimposed on the video screen to create a visually appealing presentation. The combined video ringback tone can be personalized in response to user input through the video ringback tone interface. When using the video ringback tone service, users can select their preferred weather type, set the push interval, and adjust alarm thresholds by clicking buttons or menus displayed on the screen. After receiving the user's customization request, the text stream content is updated according to the user's selection, generating user-customized content that meets the user's individual needs. User-customized content may include at least one of the following: customized weather type (e.g., if the user only cares about whether it rains, they can subscribe to rain warnings only), push interval (e.g., the user wants to receive a weather forecast at 8 a.m. every day), and alarm threshold (e.g., the user sets an alarm when the temperature exceeds 35°C).

[0034] This embodiment enables the presentation of weather forecast information as video ringback tones. Taking advantage of the proactive push of video ringback tones and the fact that no application needs to be installed, the weather forecast information is delivered directly to the user's terminal. At the same time, it supports user personalization and improves the convenience and user experience of the weather forecast service.

[0035] S105 generates a video ringback tone based on weather forecast information and sends the video ringback tone to the target terminal.

[0036] In the embodiments of this disclosure, video ringback tones are distributed to target terminals. Target terminals refer to user communication devices that receive video ringback tones, such as mobile phones, tablets, smartwatches, and other terminal devices that support video ringback tone functionality. During the distribution process, various distribution strategies can be adopted based on the network conditions, terminal capabilities, and service requirements of the target terminal. The first distribution strategy is proactive push: when a new weather forecast video ringback tone is generated, the video ringback tone content is proactively pushed to the target terminal, allowing the user to receive the latest weather forecast information without any operation. This strategy is suitable for daily weather forecast push services. The second distribution strategy is cached distribution: the video ringback tone content is pre-cached at network edge nodes (such as communication base stations or edge computing). When a user initiates a call, the edge node directly sends the cached video ringback tone content to the user, reducing the latency and bandwidth consumption of origin requests. This strategy is suitable for scenarios with high video ringback tone content traffic. The third distribution strategy is broadcast distribution: When weather forecast information meets preset weather conditions (such as typhoons, heavy rain, blizzards, cold waves, high temperatures, hail, and other weather conditions that may threaten public safety), the video ringback tone is proactively broadcast to all target terminals in the affected area, and repeated until a confirmation receipt is received from the user terminal. Broadcast distribution can send warning information to a large number of users at once, while repeated distribution ensures that users do not miss the warning information due to various reasons (such as brief signal interruptions, terminal muting, etc.). After receiving the video ringback tone, the user terminal can play the weather forecast video content on the screen, displaying weather conditions, warning information, and related text prompts. Users can also interact with the video ringback tone interface through buttons, such as confirming reading, viewing detailed weather information, and setting receiving preferences. Subsequent distribution strategies can be adjusted based on user interactions, such as stopping repeated distribution to users who have confirmed reading to avoid interference. Through this step, this application enables the use of video ringback tones as a medium for pushing weather forecast information, achieving proactive, timely, and widespread coverage of meteorological information. This is particularly suitable for elderly people who are not adept at using smartphone applications, and for rapidly disseminating early warning information to the public during natural disasters. Furthermore, video ringback tones do not require users to install any applications or run any processes in the background, reducing resource consumption on terminal devices and improving the universality and convenience of weather forecast services.

[0037] This embodiment provides a weather forecasting method based on video ringback tones. The method involves acquiring weather observation data from multiple meteorological data sources, performing data fusion processing on the weather observation data to obtain standard weather observation data, calculating the weights of each weather indicator in the standard weather observation data based on historical weather data, optimizing the standard weather observation data according to the weights of each weather indicator to obtain optimized weather observation data, acquiring current weather observation data, and using a sliding window technique to perform time series similarity matching on the current weather observation data and the historical weather data to determine historical weather data segments that match the current weather observation data, generating weather forecast information based on subsequent weather data corresponding to the historical weather data segments, generating a video ringback tone based on the weather forecast information, and sending the video ringback tone to the target terminal. This enables cross-platform fusion and accurate forecasting of multi-source meteorological data, solving the problem of insufficient forecast accuracy caused by the independence of weather platforms and lack of data fusion in existing technologies. At the same time, by using video ringback tones as the carrier for pushing weather forecast information, weather forecast information is delivered directly to the user's terminal without the need for the user to install any application or actively query it. This solves the problems of access barriers and insufficient user coverage caused by the need for users to actively query or install specific applications in existing technologies.

[0038] In some possible implementations, the weights of each weather indicator in the standard weather observation data are calculated based on historical weather data, and the standard weather observation data is optimized according to the weights of each weather indicator to obtain optimized weather observation data, including: Calculate the error between the observed values ​​of each weather indicator in the standard weather observation data and the corresponding weather indicators in the historical weather data, and calculate the mean square error based on the error; The weights of each weather indicator in standard weather observation data are calculated based on the mean square error. The standard weather observation data is weighted and optimized based on the weights of each weather indicator to obtain optimized weather observation data.

[0039] In this embodiment of the disclosure, firstly, the error between the observed values ​​of each weather indicator in the standard weather observation data and the corresponding values ​​in the historical weather data is calculated. Specifically, for each weather indicator among temperature, humidity, wind speed, wind direction, and air pressure, the observed value of that indicator in the standard weather observation data is subtracted from the baseline value at the same time and location in the historical weather data to obtain the error value e. The error value can be positive or negative, reflecting the direction and degree of deviation between the observed value and the baseline value.

[0040] Secondly, the mean squared error is calculated based on the errors. The error value at each time point is squared to obtain the squared error, eliminating the influence of negative values ​​and amplifying the weight of larger errors. Then, the average of the squared errors at all time points is calculated to obtain the mean squared error. The mean squared error is a non-negative number; the smaller the value, the closer the observed value of the weather indicator is to the historical baseline value; the larger the value, the greater the deviation between the observed value and the baseline value.

[0041] Next, the weights of each weather indicator in the standard weather observation data are calculated based on the mean squared error. The reciprocal of the mean squared error is taken to obtain the initial weights. The smaller the mean squared error, the larger the initial weight; conversely, the larger the mean squared error, the smaller the initial weight. To prevent division by zero errors, a very small positive number can be added to the denominator.

[0042] Finally, the standard weather observation data is weighted and optimized according to the weights of each weather indicator to obtain optimized weather observation data. The values ​​of each weather indicator in the standard weather observation data are multiplied by their corresponding weights, or a weighted average of observations from different data sources is calculated based on these weights, ensuring that weather indicators with high reliability play a more significant role in subsequent use.

[0043] This embodiment enables refined quality assessment and weighted optimization of standard weather observation data, effectively improving the accuracy of weather observation data.

[0044] In some possible implementations, a sliding window technique is used to perform time-series similarity matching on current weather observation data to determine historical weather data segments that match the current weather observation data, including: Construct a time series matrix; The sliding window is moved across the time series matrix at predetermined steps; After each swipe, calculate the similarity matrix between the current weather observation data and historical weather data within the current window; Record the window position information and similarity matrix obtained from each slide, and generate a similarity set; The similarity set is filtered based on a preset threshold to obtain a set of candidate similar segments; Increase the size of the sliding window and iterate until the number of elements in the candidate similar fragment set is less than or equal to a preset threshold. Then, determine the historical weather data fragments corresponding to the elements in the candidate similar fragment set as historical weather data fragments that match the current weather observation data.

[0045] In this embodiment, a sliding window technique is used to perform time-series similarity matching on current weather observation data to determine matching historical weather data segments. First, a time-series matrix is ​​constructed, arranging historical weather data in chronological order. Second, a sliding window of size k is defined and slides across the time-series matrix at predetermined steps. After each slide, the similarity matrix between the current weather observation data and historical weather data within the current window is calculated. Euclidean distance is used for similarity calculation, taking into account differences in indicators such as temperature, humidity, wind speed, wind direction, and air pressure. Then, the window position information and similarity matrix obtained from each slide are recorded to generate a similarity set. The similarity set is filtered based on a preset threshold, and elements with high similarity are stored in a candidate similar segment set. Finally, the sliding window size is increased and the calculation is iteratively performed until the number of elements in the candidate similar segment set is less than or equal to a preset threshold. The historical weather data segment corresponding to that element is then determined as a matching segment for predicting future weather.

[0046] In some possible implementations, video ringback tones are generated based on weather forecast information, including: Based on the weather type in the weather forecast information, select the corresponding video template from the preset video template library; Dynamically generate text stream content based on weather forecast information; The text stream content is combined with the video template in real time to generate a video ringback tone; In response to the user's customized operation input through the video ringback tone interface, the text stream content is updated and user-customized content is generated; the user-customized content includes at least one of customized weather type, push interval, and alarm threshold.

[0047] In this embodiment, a video ringback tone is generated based on weather forecast information. First, a corresponding video template is selected from a preset video template library based on the weather type in the weather forecast information. Weather types include sunny, rainy, cloudy, overcast, foggy, sandstorm, typhoon, high temperature, hail, and cold wave. Second, text stream content is dynamically generated based on the weather forecast information, converting key data such as temperature, precipitation probability, and wind speed into natural language text. Then, the text stream content and the video template are synthesized in real time to generate a video ringback tone, with the video layer and text layer controlled independently. Finally, in response to a user's customized operation input through the video ringback tone interface, the text stream content is updated to generate user-customized content, which includes at least one of customized weather type, push interval, and alarm threshold.

[0048] In some possible implementations, it also includes: When the weather forecast information meets the preset weather conditions, the video ringback tone is actively sent to the target terminals in the affected area via broadcast, and the message is sent repeatedly until a confirmation receipt is received.

[0049] In this embodiment, when weather forecast information meets preset weather conditions, a video ringback tone is proactively broadcast to target terminals within the affected area, and this broadcast is repeated until a confirmation receipt is received. Preset weather conditions include severe weather that may threaten public safety, such as typhoons, heavy rain, blizzards, cold waves, high temperatures, hail, sandstorms, and dense fog. Broadcasting is a one-to-many communication method, allowing the video ringback tone to be sent to all target terminals within a designated area at once. Repeated broadcasting ensures that users do not miss warning information due to signal interruption or a muted terminal. Broadcasting stops upon receiving a confirmation receipt; otherwise, it is repeated at preset intervals.

[0050] In some possible implementations, it also includes: Obtain the family number information set by the user and push the weather forecast information to the target terminal corresponding to the family number.

[0051] In this embodiment, the system obtains the family number information set by the user and pushes the weather forecast information to the target terminal corresponding to the family number. The family number is a communication number designated by the user for other users who have a kinship or close relationship with the user; the user can set one or more family numbers. After obtaining the family number information, it is stored in the user's account. After generating the weather forecast information, in addition to pushing it to the user, the weather forecast information is also pushed to designated terminals according to the family number settings. The pushed content can be personalized according to the family number, such as focusing on temperature changes, precipitation, etc., making it convenient for users to pay attention to the weather conditions in their family's area and promptly remind them to take precautions.

[0052] In some possible implementations, weather observation data from multiple meteorological data sources are acquired, and the weather observation data is fused to obtain standard weather observation data, including: Weather observation data shall be obtained from at least two of the following sources: meteorological bureau database, meteorological satellite data, meteorological radar data, meteorological observation station data, environmental monitoring station data, early warning information release center, geographic location data platform, and civilian meteorological observation data. Weather observation data is processed into a multidimensional data matrix containing time and location identifiers; the row vectors or column vectors of the multidimensional data matrix represent at least one of the following weather indicators at different time points: temperature, humidity, wind speed, wind direction, and air pressure.

[0053] In this embodiment, weather observation data is acquired from multiple meteorological data sources, including at least two of the following: meteorological bureau databases, meteorological satellite data, meteorological radar data, meteorological observation station data, environmental monitoring station data, early warning information release centers, geographic location data platforms, and civilian meteorological observation data. The weather observation data is processed into a multidimensional data matrix containing time and location identifiers. The row vectors or column vectors of the multidimensional data matrix represent at least one weather indicator among temperature, humidity, wind speed, wind direction, and air pressure at different time points. Through data cleaning, time alignment, spatial alignment, and format conversion, the heterogeneous data is integrated into standardized weather observation data in a unified format, providing high-quality input for subsequent weight calculations and weather forecasting.

[0054] To make the method provided in this disclosure clearer, a weather forecasting method based on video ringback tones will be described in detail below with reference to the following examples, such as... Figure 2 As shown, the following processes are included: S201: Collect weather observation data from multiple meteorological data sources and perform data fusion processing.

[0055] For example, multiple meteorological data sources can be implemented as follows: meteorological bureau databases, meteorological satellite data, meteorological radar data, meteorological observation station data, environmental monitoring station data, national early warning information release center data, global geographic location data platform, provincial / municipal / county-level local meteorological bureau data, and observation data from individual meteorological enthusiasts. By accessing as many of these meteorological data sources as possible, the raw data is aggregated and sent to the meteorological ringback tone cloud platform.

[0056] In the data fusion processing stage, since the data comes from multiple platforms and has different formats, data fusion technology is used to organize the data from various sources into target data and unify the data format. The unified data format includes time, location, and at least one of the following weather indicators: temperature, humidity, wind speed, wind direction, and air pressure. The data is numbered forward along the time axis. The data is constructed into a multi-dimensional data matrix containing time and location identifiers, where each column represents all features (temperature, humidity, wind speed, wind direction, air pressure) at a given time point, and each row represents the value of a feature at different time points. After fusion processing, standard weather observation data is obtained.

[0057] S202: Calculate the weights of each weather indicator in the standard weather observation data and perform weighted optimization.

[0058] For example, using historical weather data released by the China Meteorological Administration (as an example of historical weather data) as a benchmark, the weights of each weather indicator in the standard weather observation data are calculated. Specifically: First, the error between the observed values ​​of each weather indicator in the standard weather observation data and the corresponding values ​​in the historical weather data is calculated. To amplify the error and eliminate negative values, the squared error is calculated. Then, the mean squared error of the error at all time points is calculated. The mean squared error of each weather indicator is inverted to obtain its weight. A smaller mean squared error corresponds to a larger weight, indicating that the observed value of that weather indicator is more reliable.

[0059] Then, the standard weather observation data is weighted and optimized based on the calculated weights of each weather indicator. For example, for temperature, the temperature values ​​in the standard weather observation data are adjusted according to their weights; the same applies to other weather indicators such as humidity and wind speed. After weighted optimization, optimized weather observation data is obtained.

[0060] S203: Use a sliding window to match historical weather data to predict future weather.

[0061] For example, weather forecasting is performed using time series analysis methods, leveraging optimized weather observation data and historical weather data. Specifically: Construct a time series matrix. Construct a sliding window. Define a sliding window of size k, with one inlet and one outlet. The sliding direction is from the historical starting point to the most recent time point. Calculate similarity. Use Euclidean distance to measure the similarity between current weather observation data and historical weather data within the sliding window. Record the sliding window results. Define two sets. and Each element in the set contains Each time the window slides, the window similarity matrix D is calculated, carrying the entrance and exit addresses of the window. Each element is then used... This indicates that the data is saved to set S. The window slides repeatedly until the time series calculation is complete. Threshold filtering is then applied. A percentage is used to set the threshold parameter, initially set at... Set the value to be less than or equal to 99% of the elements in set S (that is, sort all elements in set S in ascending order and take the value of the 1st percentile element). (less than or equal to) The elements are stored in a new set. In the middle, simultaneously clear the set Iterative calculation. Increase the window size and repeat the above steps. Set The similarity of each element in the matrix is ​​recalculated to obtain a new similarity matrix, and the new similarity matrix is ​​saved into the set. In, until the set The number of elements in the set is less than or equal to a preset threshold (e.g., 1 or 3). In this case, the set will be... The historical weather data segment corresponding to the element is identified as the historical weather data segment that matches the current weather observation data. Subsequent weather data from this matched historical weather data segment are then used to predict future weather and generate weather forecast information.

[0062] S204: Generate video ringback tone.

[0063] For example, generating video ringback tones based on weather forecast information specifically includes: Template selection: Based on the weather type in the weather forecast information (such as sunny, rainy, cloudy, overcast, fog, sandstorm, strong wind, typhoon, tornado, high temperature, hail, cold wave, etc.), select the corresponding video template from the preset video template library.

[0064] Audio accompaniment: Match background sounds related to the weather (such as rain, wind, birdsong, etc.).

[0065] Text stream generation: Dynamically generate text stream content based on weather forecast information.

[0066] Real-time synthesis: Text stream content is synthesized with video templates in real time to generate video ringback tones. The video ringback tones adopt a layered structure, including a video layer and an interactive layer (text layer). The video layer and the interactive layer are controlled independently and can be updated and transmitted independently.

[0067] User interaction: Users select customized content through the video ringback tone interface. In response to the user's customization action, the system updates the text options based on the user's selection, generating the customized content. User-customized content includes at least one of the following: customized weather type, push interval, and alarm threshold.

[0068] S205: Send video ringback tones to the target terminal.

[0069] For example, a video ringback tone can be sent to the target terminal. The sending methods include: Detect the network type of the user terminal and select an appropriate delivery strategy based on the network conditions.

[0070] Video ringback tones can be pre-cached at base stations in various locations (as an example of network edge nodes). By utilizing edge computing technology, computing and storage capabilities are distributed to network edge nodes close to users, providing low-latency and efficient local caching services.

[0071] In addition, when the weather forecast information meets the preset weather conditions (such as typhoons, rainstorms, blizzards, cold waves, high temperatures, hail and other extreme weather), the video ringback tone is actively sent to the target terminals in the affected area via broadcast, and is sent repeatedly until a confirmation receipt is received, to ensure that users can receive important weather warning information in a timely manner.

[0072] It also obtains the family number information set by the user and pushes the weather forecast information to the target terminal corresponding to the family number, making it convenient for users to pay attention to the weather safety of their family members.

[0073] To make the weather forecasting method based on video ringback tones provided in this disclosure clearer, the following is combined with... Figure 3 Please provide an explanation. For example... Figure 3 As shown, the specific steps are as follows: 1. Data collection and processing.

[0074] Data Collection: To improve the accuracy of weather forecasts, this embodiment accesses a large number of meteorological bureau databases, such as... Figure 4 As shown, through extensive data analysis, data sources include the National Meteorological Administration database, meteorological satellite data, meteorological radar data, meteorological observation station data, China National Environmental Monitoring Center, National Early Warning Information Release Center, global geographic location data platform, provincial local meteorological bureaus, municipal meteorological bureaus, county meteorological bureaus, and as many other meteorological bureaus as possible, including individual meteorological enthusiasts. The data is collected and uploaded to the meteorological ringback tone cloud platform.

[0075] Data Processing: Since the data originates from multiple platforms and has different formats, it is necessary to analyze the data formats and use data fusion technology to organize the data from various sources into the target data, unifying the data format. The data format includes time, location (temperature, etc.). ,humidity Wind speed ,wind direction air pressure Because the time points are constantly increasing, this proposal is numbered forward along the timeline. Indicates the starting point of history. Indicates the latest point in time. As time goes by, it gets bigger and bigger. Move to the right on the timeline.

[0076] The data format is as follows: each column represents the temperature, humidity, wind speed, wind direction, air pressure, etc., at each time point; each row represents the number of time points; the data in the table represents the data for the same location, as shown in Table 1.

[0077] Table 1: Data Format Table 2. Build a model to optimize the data.

[0078] Using corrected data from the China Meteorological Administration as standard data, a weather training model was established. The model was trained using corrected historical data. After training, the weight values ​​of each indicator from each meteorological bureau at each location were obtained. Based on the weights from each meteorological bureau, more accurate indicator values ​​were calculated. The specific implementation method is as follows: 1) Calculation error ( For each weather station, each indicator (location, temperature, humidity, wind speed, wind direction, air pressure) is calculated separately at time point t. The error between the observed value of each indicator at each weather station and the corrected data from the China Meteorological Administration is calculated.

[0079] in, It is the first Error at each point in time These are observed values. This is the corrected value.

[0080] 2) Calculate the squared error ( To amplify the error and eliminate negative values, at each time point... and each observation value Squaring the error:

[0081] 3) Calculate the mean square error (MSE) ): Calculate the mean square error of the error at all time points.

[0082] in, It represents the total number of points in time.

[0083] 4) Calculate weights: Calculate the weights of each indicator data. Reverse to obtain weights, smaller ones It will correspond to a larger weight.

[0084]

[0085] in, For weights.

[0086] 5) Weight Normalization: The weights are normalized so that the sum of all weights is 1, avoiding calculation errors.

[0087] 6) Constructing the weighted loss function: Construct the weighted loss function using normalized weights:

[0088] 3. Train the model using historical data.

[0089] 1) The weight of each indicator data point is used as input for model training. The above weighted loss function is used for model training. During the training process, the model will adjust its parameters to minimize this weighted loss function. The final model parameters are those that minimize the weighted mean square error, improving the model's performance under different weight conditions and making predictions more accurate on important samples.

[0090] 2) By calculating the overall weight using the weighted average formula and adjusting its influence in the average based on the weight, a more accurate overall weight for the indicator value can be obtained. :

[0091] 3) Repeat the above steps to calculate the weight of each weather indicator such as temperature, humidity, wind speed, wind direction, and air pressure for each meteorological bureau. Select the meteorological bureau with the largest weight for each weather indicator as the source of that weather indicator to obtain more accurate weather indicator data and complete the optimization of the weather indicator data.

[0092] 4. Match historical data.

[0093] By combining existing, more accurate weather data with corrected historical data, and employing time series analysis methods, including a sliding window technique, a time series window is constructed from the current weather index data. By sliding the time series window, the similarity of time segments within the window is calculated and recorded. This similarity value is then recorded until the end of the window is reached, resulting in a set of similarity scores. Then, the threshold is calculated by filtering the set by percentage. By filtering less than or equal to The elements are used to obtain another set. Increase the window size and iterate through it. The similarity of the elements in the array is recalculated until... When a set contains only one element, the time series segment corresponding to that element represents the historical weather data with the highest similarity. This historical weather segment is used to predict future weather patterns. For example... Figure 5 As shown, the details are as follows: 1) Constructing time series: Constructing a one-dimensional matrix: Each time point is treated as an element of a one-dimensional matrix, constructing a one-dimensional time series matrix. The following one-dimensional matrix is ​​constructed. :

[0094] Constructing a two-dimensional matrix: Multiple variables at each time point are treated as a vector. These vectors are arranged in chronological order to form a two-dimensional matrix, where each column represents all features at a given time point (temperature). ,humidity Wind speed ,wind direction air pressure Each row represents the value of a feature at different time points. Construct the following two-dimensional matrix. :

[0095] 2) Construct a sliding window: like Figure 6 , Figure 7 As shown, based on the above data, a window is defined. This window slides over the time series, and its content is updated each time the window is slided by adding new data and deleting old data. The following content is defined: Window size: The size of the sliding window is Window extent: A window has one entrance. and an exit The window range is Slide step size: The distance the window slides each time is... Sliding direction: The goal of this proposal is to predict the future step by step using known past data, so the sliding direction starts from the historical starting point. To the latest time point .

[0096] 3) Calculate similarity: Euclidean distance is used to measure the similarity between current data and historical data on a sliding window, at the first time point. and the second time point The similarity between two time points is calculated as follows:

[0097] for At each time point, a window similarity matrix can be calculated. :

[0098] in, It is the size of the sliding window. It is the entry point for a time segment. It is a time segment exit.

[0099] 4) Record the results of the sliding window: Define two sets and Each element in the set contains , It is the address at the window entry point. It is the address of the window exit. It is the similarity matrix of the windows.

[0100] Each time the window slides, the data within the window is processed according to step 3 to calculate the window similarity matrix. It also carries the entrance and exit addresses of the window, and each element uses... Represent and save into a collection The window slides repeatedly until the time series calculation is complete, and all data is saved into a set. In, such as Figure 8 As shown.

[0101] 5) Threshold filtering: Step 4 above yields a set We use percentages to set the threshold parameter, the initial threshold. Set as less than or equal to set The values ​​of 99% of the elements, that is, the set Sort all elements in ascending order, and take the value of the 1st percentile element. Less than or equal to The elements are stored in a new set. In the middle. Simultaneously clear the set. .

[0102] 6) Iterative calculation: Increase the window size and repeat the above steps to collect. The similarity of each element in the matrix is ​​recalculated to obtain a new similarity matrix, and the new similarity matrix is ​​saved into a set. In, until the set Given only one element, we obtain the most similar fragment, and use the future weather of the most similar fragment to determine the future weather at the present time.

[0103] 5. Video ringback tone creation.

[0104] This embodiment proposes a novel video ringback tone interaction method. The video ringback tone consists of two parts: a video layer and an interaction layer. The transmission of both video and ringback tone content requires high bandwidth. Using 3G / 4G / 5G networks can provide sufficient bandwidth to support high-definition video streams. When network resources are limited, the video ringback tone component is cached to the mobile phone via multiple base stations, while the 2G network can transmit text interaction data in real time. The video ringback tone component employs intelligent automated video production, automatically generating corresponding video content based on weather conditions, such as... Figure 9 As shown, a simplified description is as follows: 1) Template Design: Set different types of video templates according to different types of weather, such as sunny, rainy, cloudy, overcast, fog, sandstorm, strong wind, typhoon, tornado, high temperature, hail, cold wave, etc.

[0105] 2) Audio accompaniment: Match background sounds related to the weather (such as rain, wind, birdsong).

[0106] 3) Text Stream Generation Module: Dynamically generates text content based on user needs.

[0107] 4) Synthesis Module: Combines text streams and video streams in real time to form the final ringback tone video, with layered control to allow different streams to be updated independently.

[0108] 5) User interaction module: Users select customized content through the video interface, and the text options are updated according to the user's selection.

[0109] 6. Video ringback tone distribution.

[0110] Video ringback tones are pre-cached at base stations in various locations. We utilize edge computing technology to distribute computing and storage capabilities to base stations or small data centers located at the network edge nodes close to users, providing low-latency and efficient local caching services. There are three main methods for delivering video ringback tones and weather forecasts: 1) Video ringback tone advertising ringback tone: The video ringback tone is actively sent to users through local broadcast. The text content generated by the base station in real time is superimposed on the cached video as subtitles. Users can customize the video ringback tone through simple button options on the upper layer of the video ringback tone.

[0111] 2) User-customized content delivery: This feature allows for targeted delivery of user-defined content, including customized weather types, time intervals, and alarm thresholds. Additionally, users can set up family contact numbers to receive timely weather updates.

[0112] 3) Extreme weather forecast distribution: When extreme weather is predicted, the video ringback tone will be proactively broadcast to users in the affected area, and repeated broadcasts will be used to ensure that users receive the information.

[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0114] According to embodiments of this disclosure, this disclosure also provides a weather forecast device based on video ringback tones. For example, Figure 10 This is a schematic diagram of a weather forecast device based on video ringback tones, provided in an embodiment of this disclosure. The weather forecast device 1000 based on video ringback tones includes: The acquisition module 1010 is used to acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data. The optimization module 1020 is used to calculate the weight of each weather indicator in the standard weather observation data based on historical weather data, and optimize the standard weather observation data according to the weight of each weather indicator to obtain optimized weather observation data. Matching module 1030 is used to acquire current weather observation data, and based on the optimized weather observation data and the historical weather data, to perform time series similarity matching on the current weather observation data using sliding window technology to determine historical weather data segments that match the current weather observation data. The generation module 1040 is used to generate weather forecast information based on the subsequent weather data corresponding to the historical weather data segment; The delivery module 1050 is used to generate a video ringback tone based on the weather forecast information and deliver the video ringback tone to the target terminal.

[0115] The specific implementation methods and technical effects of each module in this embodiment are similar to those in the above-described method embodiments, and will not be repeated here.

[0116] Embodiments of this disclosure also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0117] Embodiments of this disclosure also provide a computer-readable storage medium storing a computer program configured to perform the steps in any of the above method embodiments when executed.

[0118] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0119] Embodiments of this disclosure also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0120] Embodiments of this disclosure also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0121] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in 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 this disclosure.

[0122] The method provided in this disclosure has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this disclosure. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this disclosure without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this disclosure.

Claims

1. A weather forecasting method based on video ringback tones, characterized in that, include: Acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data; The weights of each weather indicator in the standard weather observation data are calculated based on historical weather data, and the standard weather observation data are optimized according to the weights of each weather indicator to obtain optimized weather observation data. Acquire current weather observation data, and based on the optimized weather observation data and the historical weather data, use sliding window technology to perform time series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data. Weather forecast information is generated based on subsequent weather data corresponding to the historical weather data segments. A video ringback tone is generated based on the weather forecast information, and the video ringback tone is sent to the target terminal.

2. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, The step of calculating the weights of each weather indicator in the standard weather observation data based on historical weather data, and optimizing the standard weather observation data according to the weights of each weather indicator to obtain optimized weather observation data includes: Calculate the error between the observed values ​​of each weather indicator in the standard weather observation data and the corresponding weather indicators in the historical weather data, and calculate the mean square error based on the error; The weights of each weather indicator in the standard weather observation data are calculated based on the mean square error. The standard weather observation data is weighted and optimized according to the weights of each weather indicator to obtain optimized weather observation data.

3. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, The step of using sliding window technology to perform time series similarity matching on the current weather observation data to determine historical weather data segments that match the current weather observation data includes: Construct a time series matrix; The sliding window is moved across the time series matrix at predetermined steps; After each swipe, calculate the similarity matrix between the current weather observation data and historical weather data within the current window; Record the window position information and similarity matrix obtained from each slide, and generate a similarity set; The similarity set is filtered based on a preset threshold to obtain a candidate similar segment set; The size of the sliding window is increased and the calculation is iterated until the number of elements in the candidate similar segment set is less than or equal to a preset threshold. The historical weather data segment corresponding to the element in the candidate similar segment set is determined as the historical weather data segment that matches the current weather observation data.

4. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, The step of generating a video ringback tone based on the weather forecast information includes: Based on the weather type in the weather forecast information, select the corresponding video template from the preset video template library; Dynamically generate text stream content based on the weather forecast information; The text stream content is combined with the video template in real time to generate a video ringback tone; In response to a user's customized operation input through the video ringback tone interface, the text stream content is updated, and user-customized content is generated; the user-customized content includes at least one of customized weather type, push interval, and alarm threshold.

5. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, Also includes: When the weather forecast information meets the preset weather conditions, the video ringback tone is actively broadcast to the target terminals in the affected area, and the broadcast is repeated until a confirmation receipt is received.

6. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, Also includes: Obtain the family number information set by the user, and push the weather forecast information to the target terminal corresponding to the family number.

7. The weather forecasting method based on video ringback tones according to claim 1, characterized in that, The process involves acquiring weather observation data from multiple meteorological data sources, performing data fusion processing on the weather observation data to obtain standard weather observation data, including: Weather observation data shall be obtained from at least two of the following sources: meteorological bureau database, meteorological satellite data, meteorological radar data, meteorological observation station data, environmental monitoring station data, early warning information release center, geographic location data platform, and civilian meteorological observation data. The weather observation data is processed into a multidimensional data matrix containing time and location identifiers; the row vectors or column vectors of the multidimensional data matrix represent at least one weather indicator among temperature, humidity, wind speed, wind direction, and air pressure at different time points.

8. A weather forecast device based on video ringback tones, characterized in that, include: The acquisition module is used to acquire weather observation data from multiple meteorological data sources, perform data fusion processing on the weather observation data, and obtain standard weather observation data. The optimization module is used to calculate the weight of each weather indicator in the standard weather observation data based on historical weather data, and optimize the standard weather observation data according to the weight of each weather indicator to obtain optimized weather observation data. The matching module is used to acquire current weather observation data, and based on the optimized weather observation data and the historical weather data, to perform time series similarity matching on the current weather observation data using sliding window technology to determine the historical weather data segment that matches the current weather observation data. The generation module is used to generate weather forecast information based on the subsequent weather data corresponding to the historical weather data segment; The delivery module is used to generate a video ringback tone based on the weather forecast information and deliver the video ringback tone to the target terminal.

9. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method described in any one of claims 1-7.

Images