A decision preference data processing method, device and equipment

Abstract

A decision preference data processing method, device and equipment, relating to the field of strategy optimization technology, including: S1, comparing and detecting a target task with a decision preference preset content template, and finding the decision preference preset content of the target task; S2, displaying a decision preference preset map corresponding to the decision preference preset content of the target task; S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map; S4, generating and saving a corresponding decision preference map according to the decision preference content of the target task; S5, generating an intelligent aggregation decision execution unit body of the target task; S6, making a decision judgment on a task that meets the characteristics of the target task through the intelligent aggregation decision unit body and giving a decision result, and executing the corresponding target decision task according to the decision result. The present invention can improve the efficiency, accuracy, consistency and economy of the decision process of the same target task by quantifying decision preferences and forming a decision model.

Description

Technical Field

[0001] The present application relates to the field of strategy optimization technology, and in particular to a decision preference data processing method, device and equipment. Background Art

[0002] With the development of the times, electronic products (such as mobile phones, tablets, computers and smart terminals, etc.) have been widely popularized and applied. The application scenarios that these electronic products can support are constantly enriched, and their functions are becoming more and more powerful. They are booming in the direction of diversification and personalization, and are gradually becoming indispensable tools for people's work and life.

[0003] Among the many rich and diverse application scenarios, a large number of tasks rely on decision support. However, with the increasing complexity of system architecture and processes, scenarios of human-machine hybrid symbiosis are becoming more and more common. In this process, a series of problems have emerged, such as insufficient transparency of decision preferences, low decision efficiency, inconsistency of decisions for the same task, and the inability of decisions to fully reflect objective laws. For this reason, how to quickly and effectively extract the decision value preferences of target tasks based on value, objectively and accurately reflect the decision value through data, and quickly obtain decision results in the same type of target tasks, this series of problems needs to be solved urgently. Summary of the invention

[0004] In order to overcome the deficiencies in the background technology, the present invention discloses a decision preference data processing method, device and equipment.

[0005] In a first aspect, the present invention provides a decision preference data processing method, comprising the following steps:

[0006] S1. Compare the detection target task with the decision preference preset content template to find the decision preference preset content of the target task;

[0007] The decision preference preset content template includes a plurality of different types of preset decision preference content;

[0008] The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags;

[0009] Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task;

[0010] S2, displaying the decision preference preset map corresponding to the decision preference preset content of the target task;

[0011] S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map;

[0012] S4. Generate and save the corresponding decision preference map according to the decision preference content of the target task;

[0013] S5, generating an intelligent aggregation decision execution unit for the target task;

[0014] S6. Make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute corresponding target decision tasks based on the decision results.

[0015] Specifically, in step S1, the matching degree between the target task and the preset content of the decision preference is calculated according to the label and the label weight, which specifically includes:

[0016]

[0017] Among them, ω j Preset the weight of content labels for decision preferences; ij is the matching value between the target task and the preset content label of the decision preference; i represents the serial number of the target task; j represents the serial number of the preset content label of the decision preference; and n represents the number of labels.

[0018] Specifically, step S1 also includes: performing semantic understanding and matching calculation on the target task description and labels through a pre-trained language model.

[0019] Specifically, step S3 includes: obtaining touch parameters of the decision preference preset map display area, and adjusting the preference parameters by touch.

[0020] Specifically, there is a mapping relationship between the touch parameter and the preference parameter, and when the touch parameter changes, the decision preference preset map changes synchronously with the preference parameter.

[0021] Specifically, step S3 also includes: there is a mapping relationship between the touch parameters and the target operation, the corresponding target operation is triggered by touching the parameters, and then the preference parameters are adjusted by touch; the target operation includes: upload operation, download operation, update operation, new operation, edit operation, delete operation, refresh operation.

[0022] Specifically, step S5 is as follows: Aggregate the attributes of the target task to output the corresponding utility, and then combine the judgment function J (Utility) to generate the intelligent aggregation decision execution unit of the target task; the aggregation function is as follows:

[0023]

[0024] Among them, Attribute is a subtask or attribute; U i (Attribute i ) is the aggregation function used for the i-th Attribute in the scenario; k i It is the weight of the i-th Attribute for the scene; Represents the aggregate of all Attributes; the judgment function J (Utility) is used for different utilities to perform different target operations.

[0025] Specifically, step S6 also includes: when the target task fails to be executed, transferring the target task to a failure queue; when the target task is successfully executed, synchronizing the decision preference content of the target task to a decision preference preset content template.

[0026] Specifically, step S4 further includes: evaluating the network status when network transmission operations are involved: calculating the comprehensive network efficiency score using a weighted summation formula, grading the network according to the comprehensive network efficiency score, and taking different operations for networks of different levels;

[0027] The formula of the network efficiency comprehensive score Score is:

[0028]

[0029] Among them, i is the index, α i is the actual value of the indicator; ω i is the weight of the indicator; σ i is the standard deviation of the indicator; i is the nonlinear index; δ i is a dynamic adjustment factor; Z is the maximum theoretical score based on all possible network states;

[0030] When Z ≥ 0.8, the network status is an excellent network, and data transmission and high-bandwidth demand tasks are fully executed;

[0031] When Z≥0.6 and Z<0.8, the network status is a good network, and balanced data synchronization is achieved;

[0032] When Z≥0.3 and Z<0.6, the network status is a poor network, at this time only the key data is updated, and the non-key data transmission is restricted; the key data includes: decision preference data adjusted by the user, task-related data input by the user; the non-key data includes: calculation process data, statistical data;

[0033] When Z<0.3, the network status is poor. At this time, data synchronization is restricted and local operations are recorded first until the network improves.

[0034] In a second aspect, the present invention further provides a decision preference data processing device, comprising the following units:

[0035] A comparison and detection unit, used for comparing and detecting the target task with the decision preference preset content template, and finding the decision preference preset content of the target task;

[0036] The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags;

[0037] Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task;

[0038] A display unit, used to display a decision preference preset map corresponding to the decision preference preset content of the target task;

[0039] An adjustment unit, used to change the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map;

[0040] A preference map generation unit, used to generate and save a corresponding decision preference map according to the decision preference content of the target task;

[0041] A decision unit generation unit, used to generate an intelligent aggregation decision execution unit for a target task;

[0042] The execution unit is used to make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute the corresponding target decision tasks according to the decision results.

[0043] In a third aspect, the present invention also provides an electronic device comprising a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program comprises instructions for executing the steps of any one of the methods described in the decision preference data processing method.

[0044] A decision preference data processing method provided by the present invention comprises the following steps: S1, comparing and detecting a target task with a decision preference preset content template, and finding the decision preference preset content of the target task; S2, displaying a decision preference preset map corresponding to the decision preference preset content of the target task; S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map; S4, generating and saving a corresponding decision preference map according to the decision preference content of the target task; S5, generating an intelligent aggregation decision execution unit body of the target task; S6, making a decision judgment on tasks that meet the characteristics of the target task through the intelligent aggregation decision unit body and giving a decision result, and executing the corresponding target decision task according to the decision result. The present invention can improve the efficiency, accuracy, consistency and economy of the decision process of the same target task by quantifying decision preferences and forming a decision model.

[0045] In addition, in the present invention, there is a mapping relationship between the touch parameters and the preference parameters. When the touch parameters change, the decision preference preset map changes synchronously with the preference parameters, making the operation more convenient and quick.

[0046] In addition, the present invention provides a method for calculating a comprehensive network efficiency score according to different network indicators, which can more accurately understand the actual operation status of the network, thereby guiding the reasonable allocation of resources, so as to perform different data operations according to different comprehensive network efficiency scores and improve the utilization rate of network resources;

[0047] In addition, the present invention can also fine-tune the weights of different indicators according to factors such as network status stability, which can make the comprehensive scoring of network efficiency more accurate and practical, and improve user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 It is a flowchart of a decision preference data processing method provided according to an embodiment of the present invention;

[0050] Figure 2 It is a demonstration diagram of a task and a queue provided according to an embodiment of the present application;

[0051] Figure 3 is a schematic diagram of a decision preference data processing device provided according to an embodiment of the present invention;

[0052] Figure 4It is a schematic diagram of a decision preference data processing device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The present invention can be explained in detail through the following embodiments. The purpose of disclosing the present invention is to protect all technical improvements within the scope of the present invention. In the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "front", "back", "left", "right", etc. indicating directions or positional relationships, they only correspond to the drawings of the present application for the convenience of describing the present invention, and do not indicate or imply that the device or element referred to must have a specific direction.

[0054] Embodiment 1

[0055] refer to Figure 1 , this embodiment provides a decision preference data processing method, including:

[0056] S1. Compare the detection target task with the decision preference preset content template to find the decision preference preset content of the target task;

[0057] The decision preference preset content template includes a plurality of different types of preset decision preference content;

[0058] The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags;

[0059] Inside the storage unit of the server, a series of extensive and common decision preference preset content templates are pre-stored. The essence of this operation is to ensure that the preset content in the server can be mapped to the electronic device side in an efficient and accurate manner, so as to effectively achieve a close synchronization between the electronic device and the server, and allow users to quickly obtain professional preference guidance. It is worth noting that even if other electronic devices log in to the server, they can also smoothly and completely synchronize and obtain this preset content. When the user has a decision preference setting requirement for the target task, he can actively and clearly send a preset data request to the server. After that, the server will compare the detection operation, find the decision preference preset content that is highly matched with the target task, and synchronize it to the client device.

[0060] A typical decision preference preset content can be a preset decision preference for a class or a specific task. In this embodiment, utility is uniformly used to quantitatively describe decision preferences. It is a subjective measure that reflects the individual's preferences and evaluations for different possible results. It is used to measure whether the goals defined by the decision maker are met. It is a clear manifestation of value and can be used as an indicator for decision-making. The size of the utility value depends on the decision maker's personal values, goals, risk attitudes and other factors. Utility is a dimensionless value ranging from 0 to 1. The closer to 1, the higher the utility, and the closer to 0, the lower the utility. The utility value represents the decision maker's preference for the attribute value. 0 means no preference and 1 means the highest point of preference.

[0061] In a typical utility map, the horizontal axis Attribute can be used to represent the specific attribute value. The attribute value can have a unit of measurement. As a measure of an individual's perception, the attribute includes definition, unit and range (from the most unacceptable to the most acceptable value interval). The attribute set constitutes the value space of the decision maker. The vertical axis represents utility. The relationship between the horizontal axis and the vertical axis can be expressed as Utility = f (Attribute). f (Attribute) is a function based on Attribute, which can be used to map Attribute to Utility. f (Attribute) can be continuous, discontinuous, linear, nonlinear, discrete or segmented. There is no limitation here. Commonly used methods include linear increase, linear decrease or table lookup. Web pages, documents, links, videos, audios, folders, apps, desktop icons, etc. can all be centered around specific utilities and attributes to further describe and characterize the target tasks and preference details.

[0062] A Utility represents a preference based on a target task. It can be judged by the judgment function J (Utility) to enter the next step of operation. J can be a threshold. Different operations can be performed according to the range of J. A Utility can perform operations of different target tasks through J (Utility).

[0063] Decision preference presets can be successfully achieved through labeling.

[0064] For the same decision preference preset content, different labels can be set reasonably and effectively, and each label gives the corresponding accurate matching value for different tasks. Under the traditional mode, the usual method is to match the labels completely without omission. However, in the current situation here, the matching method used is to combine labels with weights. After the matching operations are completed, they are sorted in order according to the degree of matching. Usually, a preset preference template that has been matched will be provided to the user. If there is a situation where multiple templates need to be provided for the user to choose, then multiple options will be presented to the user. Users can select the one that best meets their expectations and requirements from multiple templates based on their actual needs and preferences, so as to better meet the diverse needs of users in setting decision preferences.

[0065] Assume that the decision preference preset content is C, whose name can be set according to the requirements, and the label set is T = {t1, t2, ..., t n}, corresponding to the label weight set W = {ω1,ω2,...,ω n}; n is the serial number of the label, n>2; the sum of all elements in the label weight set is 1;

[0066] Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task;

[0067] Task set M = {m1,m2,...,m k}, the matching value between each task and label is V = {v i1 ,v i2 ,...,v ik}, where i = 1, 2, ..., n; k is the sequence number of the task, k> = 1;

[0068] The complete matching of the traditional model can be expressed as: if task m i With label t j If it is a complete match, the match is successful, otherwise the match fails.

[0069] In the current mode, the matching degree between the target task and the preset content of the decision preference can be expressed as: Among them, ω j Preset the weight of content labels for decision preferences; ij is the matching value between the target task and the preset content label of the decision preference; i represents the serial number of the target task; j represents the serial number of the preset content label of the decision preference; and n represents the number of labels.

[0070] After the matching degree calculation is completed, the templates are sorted according to the matching degree S. If a single preset preference template is provided, the one with the highest matching degree is selected; if multiple preset preference templates are provided, the one with the higher matching degree is selected.

[0071] In another possible implementation, when multiple targets select the same decision preference preset content template, they can also be sorted according to the degree of matching between the same type of target tasks and the decision preference preset content, and the execution order of the target tasks can be set according to the degree of matching.

[0072] In another possible implementation, the selection of the target task may be controlled by presetting a content template according to the same decision preference.

[0073] Assume that the decision preference preset content is "automobile battery selection", the label set is {range, charging time, battery cost, energy density, cycle life, safety, battery weight, warranty period, low temperature performance, fast charging capability}, the corresponding label weights are {0.15, 0.12, 0.15, 0.1, 0.1, 0.08, 0.1, 0.05, 0.1, 0.05}, and the task set is {task 1, task 2}. For example, the description of task 1 is "need a car battery with long range, short charging time, moderate cost and high safety", and the description of task 2 is "focus on the energy density and warranty period of the battery, and do not require high charging time".

[0074] Traditional complete matching method: If the task and the label are completely matched, it is successful (matching value is 1), otherwise it fails (matching value is 0). After the current pattern matching is completed, it is sorted by S. If a single template is provided, the one with the highest matching degree is selected, and if multiple templates are provided, the one with the highest matching degree is selected.

[0075] Assume that the matching values ​​of task 1 and each label are {0.8, 0.6, 0.9, 0.5, 0.7, 1, 0.4, 0.3, 0.6, 0.7}, and the matching values ​​of task 2 and each label are {0.7, 0.5, 0.8, 0.6, 0.8, 0.7, 0.5, 0.4, 0.7, 0.6}.

[0076] For Task 1, the matching degree is:

[0077] S1=0.15×0.8+0.12×0.6+0.15×0.9+0.1×0.5+0.1×0.7+0.08×1+0.1×0.4+0.05×0.3+0.1×0.6+0.05×0.7=0.687

[0078] For Task 2, the matching degree is:

[0079] S2=0.15×0.7+0.12×0.5+0.15×0.8+0.1×0.6+0.1×0.8+0.08×0.7+0.1×0.5+0.05×0.4+0.1×0.7+0.05×0.6=0.651

[0080] Sort by matching degree from large to small, with Task 1 before Task 2. If only one template is provided, Task 1 is selected; if multiple templates are provided, both Task 1 and Task 2 may be provided for the user to choose.

[0081] In another embodiment, a pre-trained language model is introduced to perform semantic understanding and matching calculation on task descriptions and labels.

[0082] In the above decision preference setting, in addition to the above matching calculation method, we can also introduce a large model algorithm to perform matching evaluation. For example, a pre-trained language model such as Llama or its subsequent versions can be used to perform semantic understanding and matching calculation on task descriptions and labels.

[0083] First, the task description and label are converted into a text format suitable for model input. Then, the model will give a comprehensive matching score based on its understanding of semantics and learned patterns. Assume that for the decision preference preset of "car battery selection", the description of task 1 is "need a car battery with long driving range, short charging time, moderate cost and high safety", and the description of task 2 is "focus on the energy density and warranty period of the battery, and do not require high charging time". These descriptions are input into the large model, and after analysis and calculation, the model gives a matching score of 0.75 for task 1 and 0.68 for task 2. Compared with previous algorithms, the large model algorithm can better handle the complexity and ambiguity of natural language, capture more semantic information, and thus provide a more accurate and comprehensive matching evaluation. Sorted by matching degree, in this case, task 1 still ranks ahead of task 2. If only one template is provided, task 1 is selected; if multiple templates are provided, both task 1 and task 2 may be provided for users to choose. It should be noted that when using large model algorithms, the models need to be properly trained and optimized to adapt to specific decision preference areas and data characteristics.

[0084] S2, displaying the decision preference preset map corresponding to the decision preference preset content of the target task;

[0085] The decision preference preset content can be set by the user or by the system default. The display form of the decision preference preset content can be mainly a decision preference map, and the preference distribution of different configurations can be displayed through various visualization methods such as scatter plots, curve graphs, heat maps or three-dimensional graphs to form a decision preference map, and it can be displayed in one or a combination of web pages, documents, links, videos, audios, folders, apps, desktop icons, etc., without limitation here.

[0086] In this embodiment, the decision preference map corresponding to the decision preference preset content is called a decision preference preset map.

[0087] In a possible implementation, the preset content is displayed by the electronic device in an online or offline state, that is, the electronic device can be connected to a server corresponding to the decision preference preset content, or the electronic device can also be not connected to a server corresponding to the decision preference preset content, such as: the electronic device can edit the preset content in an offline state. For example, the preset content can be "electric vehicle battery component selection task decision preference".

[0088] S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map;

[0089] There is a mapping relationship between the touch parameters and the preference parameters. When the touch parameters change, the decision preference preset map changes synchronously with the preference parameters.

[0090] Adjust the preference parameters of the decision preference preset content of the target task to generate the decision preference content of the target task; obtain the touch parameters of the display area of ​​the decision preference preset map, and adjust the preference parameters by touch;

[0091] The preference parameters correspond to the name, label set and label weight set in the decision preference preset content;

[0092] In a possible example of adjusting preset content, the following steps may be included:

[0093] Acquiring touch parameters for the display area of ​​the decision preference preset content, and adjusting the preference parameters of the decision preference preset content by touching;

[0094] There is a mapping relationship between the touch parameters and the target operation. The corresponding target operation is triggered by touching the parameters, and then the preference parameters are adjusted by touch; the target operations include: upload operation, download operation, update operation, new operation, edit operation, delete operation, and refresh operation.

[0095] In a specific implementation, the mapping relationship between preset touch parameters and target operations can be pre-stored in the electronic device, and then, when the user touches the preference that needs to be adjusted, the preference parameters can be adjusted through the touch display screen. The decision preference map and the decision preference parameters change synchronously according to the touch content. This step is mainly to improve the efficiency of setting decision preference parameters. Among them, for part or all of the preset content. The target operation can be at least one of the following: upload operation, download operation, update operation, new operation, edit operation, delete operation, refresh operation, etc., which are not limited here. In a specific implementation, for example, an edit operation can be performed on a preference; for another example, an upload operation can be performed on a preference; for another example, an update operation can be performed on a preference; for another example, an edit operation can be performed on a preference.

[0096] According to the mapping relationship between the preset touch parameters and the operation, the target operation corresponding to the touch parameters is determined. Wherein, in the embodiment of the present application, the touch parameters can be at least one of the following: touch force, touch time, touch position, touch trajectory, number of touches, touch frequency, touch direction, touch area, etc., which are not limited here. The touch force can be the average touch force when the touch area is within the preset area range, or the touch force can be the maximum touch force when the touch time is greater than the preset duration, etc. The preset area range and the preset duration can be set by the user or by the system default. The touch time can be a time point or a time length. For example, the touch time is the duration when the touch force is greater than the preset touch force, or the touch time can be the time point when the touch area is equal to the specified touch area, etc., wherein the preset touch force and the specified touch area can be set by the user or by the system default.

[0097] Touch parameters are data generated when a user touches the screen. These parameters can be detected by electronic devices and used to understand the user's intentions;

[0098] The electronic device pre-stores the mapping relationship between preset touch parameters and target operations. This means that when the user touches the screen in a specific way (such as long press or fast slide), the system can recognize this touch mode and map it to a specific target operation.

[0099] For example, if a user wants to edit a preference setting, they can do so by sliding or pressing the setting on the screen. After the system detects these touch parameters, it determines that the user's target operation is to edit based on the preset mapping relationship and opens the editing interface.

[0100] For example, if the user performs a long press operation, the system may recognize the corresponding target operation as a new operation, thereby allowing the user to add new preference parameters to the current decision preference preset content.

[0101] This embodiment provides a specific implementation method for performing precise adjustment, fuzzy adjustment, and transition between the two adjustment modes on a curve in a preset decision preference map by touch, so as to better improve the user experience.

[0102] 1. Precise Adjustment

[0103] Precise adjustment is mainly used in scenarios where users need to make fine and accurate adjustments to the curve.

[0104] 1. Introducing the concepts of inertia and damping

[0105] Inertia: When the user slowly and finely slides the touch screen to adjust the curve, assume that the user's touch speed is u (unit: pixel / second). The relationship between the curve change speed v (unit: curve parameter unit / second) and the user's touch speed u is expressed by a nonlinear function f precise (u) = k 11 u 3 Indicates that k 11 is the proportional coefficient for precise adjustment. When the user stops touching, the change speed of the curve will decay according to the damping. Introduce the damping coefficient d precise (Unit: 1 / second), velocity v at time t t The speed v after the time interval Δt (unit: seconds) t+Δt It can be calculated by the following formula: t+Δt =v t -d precise v t Δt. As time goes by, the velocity decreases until v approaches 0 and the curve stops changing.

[0106] Damping: Damping coefficient d precise The value of determines how fast the speed decays. precise The larger the value, the faster the speed decays and the faster the curve stops changing; d precise The smaller the value, the slower the speed decays, and the curve will continue to change for a longer time after the touch stops.

[0107] 2. Elastic feedback

[0108] When the user's touch operation makes the curve value y (unit: curve parameter unit) approach or reach the limit value y max (Unit: curve parameter unit) or y min (unit: curve parameter unit), the feedback force F (unit: Newton) generated and the distance x (unit: curve parameter unit) from the limit value are calculated through the nonlinear function g precise (x) = k 21 x 4 Indicates that k 21The feedback factor for precise adjustment.

[0109] 3. Touch accuracy and zoom

[0110] Touch accuracy: According to the user's touch speed u (unit: pixel / second) and range s (unit: pixel), the adjustment accuracy Δy (unit: curve parameter unit) of the curve is automatically adjusted. Assume that there is a nonlinear function To calculate the accuracy, where k 31 is the precision coefficient of precise adjustment. When u is small and s is small, Δy is small and fine adjustment is performed.

[0111] Scaling: by scaling factor r precise When the user zooms in on the curve, r precise >1 and usually close to 1, such as r precise =1.1 means enlarging the curve by 10%; when the user precisely reduces the curve, 0 <r precise <1 and usually close to 1, such as r precise =0.9 means shrinking the curve by 10%.

[0112] 4. Multi-touch and coordinated adjustment

[0113] Assume that there are two fingers touching at (x1, y1) and (x2, y2) (unit: pixel). The operation of one finger controls the translation of the curve, for example, the translation amount Δx = x2-x1; the operation of another finger controls the stretching of the curve, for example, the stretching coefficient

[0114] 2. Fuzzy Adjustment

[0115] Fuzzy adjustment is suitable for users to perform quick and rough curve adjustment operations.

[0116] 1. Introducing the concepts of inertia and damping

[0117] Inertia: When the user quickly and significantly slides the screen to adjust the curve, assume that the user's touch speed is u (unit: pixel / second). The relationship between the curve change speed v (unit: curve parameter unit / second) and the user's touch speed u is expressed by a nonlinear function f fuzzy (u) = k 12 u 3 Indicates that k 12 is the proportional coefficient of fuzzy adjustment. When the user stops touching, the change speed of the curve will decay according to the damping. Introduce the damping coefficient d fuzzy (Unit: 1 / second), velocity v at time t t The speed v after the time interval Δt (unit: seconds) t+Δt It can be calculated by the following formula:t+Δt =v t -d fuzzy v t Δt. As time goes by, the speed gradually decreases until v t As it approaches 0, the curve stops changing.

[0118] Damping: Damping coefficient d fuzzy The value of determines how fast the speed decays. fuzzy The larger the value, the faster the speed decays and the faster the curve stops changing; d fuzzy The smaller the value, the slower the speed decays, and the curve will continue to change for a longer time after the touch stops.

[0119] 2. Elastic feedback

[0120] When the user's touch operation makes the curve value y (unit: curve parameter unit) approach or reach the limit value y max (Unit: curve parameter unit) or y min (unit: curve parameter unit), the feedback force F (unit: Newton) generated and the distance x (unit: curve parameter unit) from the limit value are calculated through the nonlinear function g fuzzy (x) = k 22 x 3 Indicates that k 22 is the feedback coefficient of fuzzy regulation.

[0121] 3. Touch accuracy and zoom

[0122] Touch accuracy: According to the user's touch speed u (unit: pixel / second) and range s (unit: pixel), the adjustment accuracy Δy (unit: curve parameter unit) of the curve is automatically adjusted. Assume that there is a nonlinear function To calculate the accuracy, where k 32 is the precision coefficient of fuzzy adjustment. When u is large and s is wide, Δy is large and a coarser-grained adjustment is performed.

[0123] Scaling: by scaling factor r fuzzy When the user fuzzily zooms in on the curve, r fuzzy >1 and usually large, such as r fuzzy =2 means to magnify the curve twice; when the user blurs and reduces the curve, 0 <r fuzzy <1 and usually small, such as r fuzzy =0.5 means shrinking the curve to half of its original size.

[0124] 4. Multi-touch and coordinated adjustment

[0125] Assume that there are two fingers touching at (x1, y1) and (x2, y2) (unit: pixel). The operation of one finger controls the translation of the curve, for example, the translation amount Δx = x2-x1; the operation of another finger controls the stretching of the curve, for example, the stretching coefficient

[0126] 3. Transition from precise regulation to fuzzy regulation

[0127] In actual touch operations, when the user gradually increases the speed or expands the touch range from fine and slow touch, the system needs to achieve a smooth transition from precise adjustment to fuzzy adjustment.

[0128] The timing of the transition can be determined by setting a threshold. For example, when the user's touch speed u exceeds a certain value u threshold , or the touch range s exceeds a certain range s threshold When the system is adjusted, it gradually switches from the precise adjustment algorithm and parameters to the fuzzy adjustment algorithm and parameters.

[0129] During the transition process, linear interpolation or other smooth transition methods can be used to adjust the coefficients in the algorithm to achieve smooth changes in the adjustment effect. For example, if the current transition stage is in progress, the coefficient of precise adjustment is k precise , the coefficient of fuzzy adjustment is k fuzzy , the transition ratio is p (0≤p≤1), then the coefficient k actually used can be calculated by the following formula: k = 1-p)k precise +pk fuzzy .

[0130] S4. Generate and save the corresponding decision preference map according to the decision preference content of the target task;

[0131] This process can be summarized into the following key steps:

[0132] 1) Define attributes and utility functions: First, clarify the attributes to be considered in the selection and their value ranges. The conversion of each attribute to utility is a mathematical model, such as linear, nonlinear, table lookup model or deep learning model to express the law of utility changing with attribute value. This step ensures that preferences can be quantified and compared.

[0133] 2) Calculate the utility value of each attribute: Based on the given specific values, use the utility function of each attribute to calculate the utility value. For example, the utility corresponding to a range of 500 kilometers is 0.6, and the utility corresponding to a charging time of 6 hours is about 0.3, etc.

[0134] 3) Construct a decision preference map: Integrate the utility value of each attribute into a multidimensional vector, where each dimension represents the utility of an attribute. This multidimensional vector can be regarded as a point in the decision space, representing the preferred position of a specific configuration.

[0135] The preference distribution of different configurations can be displayed through various visualization methods such as scatter plots, heat maps or three-dimensional maps to form a decision preference map. The position of each point reflects its comprehensive preference level under all considerations.

[0136] 4) Data storage and management: The calculated attribute utility values ​​and preference map data are organized and prepared for storage on the server. Data storage can be done using relational databases (such as MySQL), NoSQL databases (such as MongoDB), or cloud storage services (such as Amazon S3), depending on the data size, access patterns, and subsequent analysis requirements. Ensure structured storage of data for subsequent query, analysis, and application.

[0137] 5) Application and training of deep learning models: For models such as MLP and CNN that use deep learning models to predict utility values, a large amount of relevant data needs to be collected in advance for model training. During the training process, the model parameters are adjusted to optimize the prediction effect and ensure that the model can accurately reflect the actual utility trend. After training, the model is deployed on the server so that the utility value can be updated in real time or regularly based on new data.

[0138] The decision-making task of this embodiment, when the decision is made when the server is running, or in a distributed application, data will be transmitted through the network. Therefore, when it comes to network transmission operations, such as data synchronization or data update, the network status will be evaluated in real time: the current network status represented by the network rate is obtained; among them, indicators such as network rate, delay, and packet loss rate are analyzed in real time. The weighted summation formula is used to calculate the comprehensive score of network efficiency. For each indicator i, the score calculation formula is:

[0139]

[0140] Among them, α i is the actual value of the indicator, obtained through network monitoring tools; ω i is the weight of the indicator, reflecting the relative importance, which is set according to system design and experience; i is the standard deviation of the indicator, which measures the degree of fluctuation and is obtained through statistical analysis of data; i is a nonlinear index, reflecting the nonlinear effect, and is determined based on the network characteristics; δ i To dynamically adjust the factors, the weights are fine-tuned according to factors such as network status stability.

[0141] The network efficiency comprehensive score Score is normalized to ensure that the score range is between 0 and 1. The normalization formula is:

[0142]

[0143] where Z is the maximum theoretical score based on all possible network states.

[0144] Assume that there are m indicators, and the scores of each indicator in the most ideal state are S max1 ,S max2 ,…,S maxm ,but

[0145] Rating application strategy:

[0146] When Z ≥ 0.8, the network status is an excellent network, and data transmission and high-bandwidth demand tasks are fully executed;

[0147] When Z≥0.6 and Z<0.8, the network status is a good network, and the data synchronization is balanced to ensure stability and efficiency;

[0148] When Z≥0.3 and Z<0.6, the network status is a poor network. At this time, only critical data is updated and non-critical data transmission is restricted;

[0149] When Z<0.3, the network status is poor. At this time, data synchronization is restricted and local operations are recorded first until the network improves.

[0150] The data transmitted over the network is divided into priorities. The higher the priority, the higher the criticality. Critical data refers to parameter data adjusted by the user, such as decision preference data adjusted by the user, task-related data input by the user, etc. Non-critical data refers to data used to support system operation, such as calculation process data, statistical data, etc.

[0151] The method of calculating the comprehensive network efficiency score according to different network indicators in this embodiment can more accurately understand the actual operation status of the network, thereby guiding the reasonable allocation of resources, performing different data operations according to different comprehensive network efficiency scores, and improving the utilization rate of network resources;

[0152] Assume that the three indicators of network download speed, upload speed and latency are evaluated. The indicators and related parameters are set as follows:

[0153] Download speed: actual value α1 = 50 Mbps, weight ω1 = 0.4, standard deviation σ1 = 1.51 Mbps, nonlinear index γ1 = 1.2, dynamic adjustment factor δ1 = 1.2; upload speed: actual value α2 = 20 Mbps, weight ω2 = 0.3, standard deviation σ2 = 0.8 Mbps, nonlinear index γ2 = 1.1, dynamic adjustment factor δ2 = 0.08. Delay: actual value α3 = 50 milliseconds, weight ω3 = 0.3, standard deviation σ3 = 5 milliseconds, nonlinear index γ3 = 1.0, dynamic adjustment factor δ3 = 0.05.

[0154] First, calculate the score of each indicator separately: Download speed score: (0.4+0.1×1.51)×50 1.2 ≈92.49. Upload speed score: (0.3+0.08×0.8)×20 1.1 ≈28.87. Delay score: (0.3+0.05×5)×50≈27.5.

[0155] Then calculate the overall network efficiency score Assuming that under the most ideal conditions, the maximum download speed score is S max1 =100, maximum upload speed score S max2 =50, maximum delay score S max3 =30, then Z = 100 + 50 + 30 = 180, This network condition is a good network (0.6-0.8), and data synchronization should be balanced to ensure stability and efficiency.

[0156] step:

[0157] Set initial weights: According to system design and experience, set initial weights for relevant indicators. For example, the download speed weight ω_1 is 0.4, the upload speed weight ω_2 is 0.3, and the delay weight ω_3 is 0.3.

[0158] Continuously monitor network status: Use network monitoring tools to obtain real-time data such as hourly download speed, upload speed, and latency.

[0159] Stability assessment: Based on the data monitored during the hour, the stability of the network is assessed. For example, the standard deviation σ_1 of the actual value of the download speed α_1, the average value and standard deviation of the upload speed, and the average value and standard deviation of the latency are calculated during the hour.

[0160] Determine the adjustment direction:

[0161] If the stability of the download speed is good (small standard deviation), for example, σ_1<a certain threshold, the dynamic adjustment factor δ_1 can be appropriately increased, so that the weight ω_1 of the download speed is fine-tuned to increase. If the stability of the upload speed deteriorates (the standard deviation increases), for example, σ_2>a certain threshold, the dynamic adjustment factor δ_2 can be appropriately reduced, resulting in a fine-tuned decrease in the weight ω_2 of the upload speed. If the stability of the delay remains at a good level (small standard deviation), for example, σ_3<a certain threshold, the dynamic adjustment factor δ_3 can be appropriately increased to increase the weight ω_3 of the delay indicator.

[0162] Consider the characteristics of indicators: In addition to stability, the characteristics of each indicator should also be considered. For example, for some services with particularly high real-time requirements, even if the delay stability is good, it may be necessary to increase its initial weight ω_3.

[0163] Real-time review and update: Re-evaluate the network status every hour and adjust the dynamic adjustment factor δ_i based on new data and conditions.

[0164] Combined with actual business needs: If the current business mainly relies on data upload, the weight ω_2 of the upload speed indicator can be increased accordingly.

[0165] Test and verify: After adjusting the weights, immediately observe the changes in the comprehensive network efficiency score and the impact on related services over the next period of time. For example, if the comprehensive network efficiency score improves after adjusting the weights and the service runs more smoothly, it means that the adjustment is effective; otherwise, you need to return to the previous steps to further optimize the adjustment strategy.

[0166] For example, in a certain hour, the average download speed is 50 Mbps, with a standard deviation of 2 Mbps; the average upload speed is 20 Mbps, with a standard deviation of 1.5 Mbps; the average latency is 50 milliseconds, with a standard deviation of 2 milliseconds. Assume that the thresholds for download speed stability are 3 Mbps, upload speed stability threshold is 2 Mbps, and latency stability threshold is 3 milliseconds.

[0167] According to these data, the download speed is relatively stable (2Mbps<3Mbps), and we can consider increasing the dynamic adjustment factor of the download speed δ_1 from the initial 1.2 to 1.3; the upload speed is also stable (1.5Mbps<2Mbps), and δ_2 remains unchanged; the latency is also stable (2ms<3ms), and δ_3 can be increased from 0.05 to 0.06.

[0168] Then, calculate the comprehensive network efficiency score according to the new weights and observe the business operation to determine whether the adjustment is effective. If the effect is not ideal, you can further adjust the dynamic adjustment factor or weight setting. In actual applications, the setting of thresholds and the specific adjustment range need to be tested and optimized multiple times according to the network environment, business needs and experience.

[0169] This embodiment fine-tunes the weights of different indicators according to factors such as network status stability, which can make the comprehensive network efficiency score more accurate and practical, and improve user experience.

[0170] The embodiment of the present application also supports offline editing, that is, it can be edited and saved in an environment without a network, involving three levels of cache: memory, disk, and database. Taking "decision preference data" as an example, using the SQLite database, large data may cause database query performance to decline, so the data size must be determined first, and when it exceeds 200K, it is stored in a file. When a decision preference data preset content is downloaded from the server to the client, it is written to the file first, and then to the database after success, and then to the memory cache. When reading, the memory cache is read first to speed up reading and reduce file IO.

[0171] S5, generating an intelligent aggregation decision execution unit for the target task;

[0172] Through aggregate functions Aggregate the attributes of the target task, and then combine the judgment function J (Utility) to generate the intelligent aggregation decision execution unit of the target task.

[0173] When a target task consists of multiple attributes or subtasks, it is necessary to aggregate the subtasks or multiple attributes to judge the target task.

[0174] Aggregate functions Indicates the corresponding utility of the aggregated output of all subtasks or attributes. Represents the aggregate of all subtasks or attributes. The aggregation function is not limited here.

[0175] Preferably, this embodiment provides the following aggregation functions:

[0176]

[0177] Among them, Attribute is a subtask or attribute; U i (Attribute i ) is the aggregation function used for the i-th Attribute in the scenario. Users can choose according to the actual situation. i Utility = f(Attribute i ) abbreviation; k iIt is the weight of the i-th Attribute for the scene;

[0178] The result calculated by the aggregation function is combined with J (Utility) to form an aggregate, which is used for the subsequent calculation of target tasks with the same characteristics. The aggregation of the target task intelligent aggregation decision execution unit is executed on the server.

[0179] A Utility represents a preference based on a target task. It can be judged by the judgment function J (Utility) to enter the next step of operation. J can be a threshold set by the decision maker. Different operations can be performed according to the range of J. A Utility can perform operations of different target tasks through J (Utility).

[0180] This embodiment provides a specific embodiment of an intelligent aggregation decision execution unit body for generating target tasks based on electric vehicle component selection, and also provides specific values ​​and performs calculations:

[0181] First, define utility to quantitatively describe decision preferences. The following are ten possible attributes, specific value ranges, and the assumed utility function form:

[0182] 1) Battery life: In kilometers, the range is from 200 kilometers to 800 kilometers. Assuming the utility function is a piecewise linear model, when the range is 200-400 kilometers, the utility value increases linearly from 0 to 0.4; when it is 400-600 kilometers, the utility value increases linearly from 0.4 to 0.8; when it is 600-800 kilometers, the utility value increases linearly from 0.8 to 1. If the actual range is 500 kilometers, the utility value is 0.6.

[0183] 2) Battery charging time: in hours, ranging from 2 hours to 10 hours. Assume the utility function is a nonlinear model In this function, the denominator "8" is determined based on the range of charging time (10-2 = 8 hours) and is used to normalize the charging time. When the charging time is 2 hours, the utility value is 1; when the charging time is 10 hours, the utility value is 0. If the charging time is 6 hours, the utility value is about 0.7.

[0184] 3) Battery cost: in RMB, ranging from 10,000 to 50,000 yuan. Assuming the utility function is a piecewise function, when the cost is 10,000-20,000 yuan, the utility value is 1; when it is 20,000-40,000 yuan, the utility value decreases linearly from 1 to 0.5; when it is 40,000-50,000 yuan, the utility value decreases linearly from 0.5 to 0. If the cost is 35,000 yuan, the utility value is about 0.375.

[0185] 4) Battery energy density: in watt-hours / kg, ranging from 100 watt-hours / kg to 300 watt-hours / kg. Assuming the utility function is a lookup table model, the utility values ​​corresponding to different energy densities are set in advance, such as 100 watt-hours / kg corresponds to 0, 200 watt-hours / kg corresponds to 0.6, and 300 watt-hours / kg corresponds to 1. If the energy density is 200 watt-hours / kg, the utility value is 0.6.

[0186] 5) Battery cycle life: in units of charge and discharge times, ranging from 500 to 2000 times. Assume the utility function is a complex model The function is constructed in logarithmic form. The principle is that as the cycle life increases, its contribution to the overall utility gradually increases, but the growth rate gradually slows down. When the cycle life is 500 times, the utility value is close to 0; when the cycle life is 2000 times, the utility value is close to 1. If the cycle life is 1200 times, the utility value is about 0.58.

[0187] 6) Battery safety: Based on a certain safety rating, for example, from level 1 to level 5. Assume that the utility value is 0 at level 1 and 1 at level 5, and the utility value increases linearly with the level. If the safety rating is level 4, the utility value is 0.8.

[0188] 7) Battery weight: in kilograms, ranging from 100 kg to 300 kg. Assuming the utility function is a piecewise linear model, when the weight is 100-150 kg, the utility value decreases linearly from 1 to 0.8; when the weight is 150-250 kg, the utility value decreases linearly from 0.8 to 0.2; when the weight is 250-300 kg, the utility value decreases linearly from 0.2 to 0. If the weight is 200 kg, the utility value is about 0.5.

[0189] 8) Battery warranty period: in years, ranging from 3 to 8 years. Assuming the utility function is a linear model, the utility value is 0 when the warranty period is 3 years, and the utility value is 1 when the warranty period is 8 years. The utility value increases linearly with the warranty period. If the warranty period is 5 years, the utility value is 0.4.

[0190] 9) Low temperature performance of the battery: The battery life attenuation rate in a specific low temperature environment is used as an indicator, ranging from 0% to 50%. A deep learning model based on a multi-layer perceptron (MLP) is constructed to calculate the utility value. The MLP model consists of an input layer with the input as the attenuation rate; two hidden layers with 64 and 32 neurons respectively, and the activation function is ReLU; and an output layer with the output as the utility value and the activation function is Sigmoid. The model is trained with training data so that it can accurately predict the utility value based on the input attenuation rate. Assuming the actual attenuation rate is 15%, the utility value output by the model is approximately 0.7.

[0191] 10) Fast charging capability of the battery: The time required to reach 80% power under specific fast charging conditions is used as an indicator, ranging from 15 minutes to 60 minutes. A deep learning model based on a convolutional neural network (CNN) is used to calculate the utility value. The CNN model consists of an input layer that receives charging time data; two convolutional layers with kernel sizes of 3 and 5, 16 and 32 channels, and ReLU activation function; a pooling layer with maximum pooling; a fully connected layer with 64 neurons and ReLU activation function; and an output layer that outputs the utility value and the activation function is Sigmoid. The model is trained with training data so that it can accurately predict the utility value based on the input charging time. If the actual charging time is 25 minutes, the utility value output by the model is approximately 0.85.

[0192] Assume that the specific parameters of a battery are: 500 km range, 6 hours charging time, 35,000 yuan cost, 200 Wh / kg energy density, 1,200 cycles, 4 safety rating, 200 kg weight, 5 years warranty, 15% low temperature performance attenuation rate, 25 minutes fast charging time. The utility values ​​of each attribute are: range: 0.6, charging time: 0.3, cost: 0.375, energy density: 0.6, cycle life: 0.58, safety: 0.8, weight: 0.5, warranty: 0.4, low temperature performance: 0.7, fast charging capability: 0.85.

[0193] Use the above formula for the following three different aggregation scenarios: Perform the calculation:

[0194] Scenario 1 (Urban Commute):

[0195] Assume that the weights of these ten attributes in the urban commuting scenario are k1 to k 10 They are 0.15, 0.12, 0.15, 0.1, 0.1, 0.08, 0.1, 0.05, 0.1, and 0.05 respectively.

[0196] The aggregate utility value U1 is:

[0197] U1=0.6×0.15+0.3×0.12+0.375×0.15+0.6×0.1+0.58×0.1+0.8×0.08+0.5×0.1+0.

[0198] 4×0.05+0.7×0.1+0.85×0.05=0.55675,

[0199] Since 0.55675<preset decision threshold in the urban commuting scenario, according to the judgment function J (Utility), an economical brand of battery components should be selected.

[0200] Scenario 2 (Long-distance travel):

[0201] Assume that the weights of these ten attributes in the long-distance travel scenario are k1 to k 10 They are 0.2, 0.1, 0.15, 0.15, 0.12, 0.08, 0.08, 0.4, 0.05, 0.1 and 0.07 respectively.

[0202] The aggregate utility value U2 is:

[0203] U2=0.6×0.2+0.3×0.1+0.375×0.15+0.6×0.15+0.58×0.12+0.8×0.08+0.5×0.08+0.

[0204] 4×0.05+0.7×0.1+0.85×0.07=0.66935,

[0205] In the long-distance travel scenario, 0.66935> the preset decision threshold, according to the judgment function J (Utility), so a medium-brand battery component should be selected.

[0206] Scenario 3 (Economy First):

[0207] Assume that the weights of these ten attributes in the economic priority scenario are k1 to k 10 They are 0.1, 0.15, 0.2, 0.08, 0.08, 0.05, 0.12, 0.15, 0.08 and 0.09 respectively.

[0208] The aggregate utility value U3 is:

[0209] U3=0.6×0.1+0.3×0.15+0.375×0.2+0.6×0.08+0.58×0.08+0.8×0.05+0.5×0.12+0.

[0210] 4×0.15+0.7×0.08+0.85×0.09=0.5669

[0211] In the economy priority scenario, 0.5669 < the preset decision threshold, according to the judgment function J (Utility), so an economical brand of battery components should be selected.

[0212] The preset decision threshold is 0.6, which is set by the decision maker based on experience.

[0213] Through the analysis of the above different scenarios, the comprehensive utility of the battery can be accurately evaluated according to specific needs and weight settings, and corresponding brand selection decisions can be made. Moreover, each attribute can be used as a decision-making unit separately. For example, if only the attribute of cruising range is considered, when the cruising range reaches more than 600 kilometers, the battery can be directly selected without considering other attributes. Similarly, separate decisions can be made for other attributes based on specific needs and importance. When new electric vehicle models need to select battery components in the future, calculations and judgments can also be made based on these attributes and functions, using the previously formed aggregates and judgment criteria to improve decision-making efficiency and accuracy, while also better meeting the needs of different users and usage scenarios.

[0214] S6. Make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute corresponding target decision tasks based on the decision results.

[0215] refer to Figure 2 In this embodiment, the preset time interval for executing the target task can be set by the user or by the system default. When the target task arrives, it can be automatically executed. In the specific implementation, different operations can correspond to different waiting queues. The electronic device can add the target task to the waiting queue corresponding to the target operation, and take the task with the first priority level in the waiting queue in the server at every preset time interval. For example, the timer can be turned on and the weight can be adjusted at the same time to achieve taking any task in the waiting queue at every preset time interval. Furthermore, the task execution status can be detected within a certain interval, and then any task can be loaded into the execution queue. Specifically, when the execution queue includes any task, it can no longer be loaded repeatedly. On the contrary, when the execution queue does not include any task, it can be added to the execution queue. When the target task is obtained, the target task can be executed to synchronize the target operation in the server, which can ensure the real-time data synchronization and avoid long-term data synchronization and affect the user experience.

[0216] Furthermore, in a possible example, the following steps may also be included:

[0217] When the target task fails to be executed, the target task is transferred to the failure queue corresponding to the target operation; or, when the target task is successfully executed, the target operation in the server is synchronized. In a specific implementation, when the target task fails to be executed, the electronic device can transfer the target task to its corresponding failure queue; conversely, when the target task is successfully executed, the target operation in the synchronization server can be realized, that is, the server completes the target operation, thereby achieving the purpose of data synchronization.

[0218] The present embodiment provides a method for processing decision preference data, comprising the following steps: S1, comparing the detection target task with the decision preference preset content template, and finding the decision preference preset content of the target task; S2, displaying the decision preference preset map corresponding to the decision preference preset content of the target task; S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map; S4, generating and saving the corresponding decision preference map according to the decision preference content of the target task; S5, generating an intelligent aggregation decision execution unit body of the target task; S6, making decision judgments and giving decision results on tasks that meet the characteristics of the target task through the intelligent aggregation decision unit body, and executing the corresponding target decision task according to the decision results. The present embodiment can improve the efficiency, accuracy, consistency and economy of the decision process of the same target task by quantifying decision preferences and forming a decision model.

[0219] Embodiment 2

[0220] refer to Figure 3 This embodiment discloses a decision preference data processing device. The present invention also provides a decision preference data processing device, including the following units:

[0221] A comparison and detection unit, used for comparing and detecting the target task with the decision preference preset content template, and finding the decision preference preset content of the target task;

[0222] The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags;

[0223] Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task;

[0224] A display unit, used to display a decision preference preset map corresponding to the decision preference preset content of the target task;

[0225] An adjustment unit, used to change the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map;

[0226] A preference map generation unit, used to generate and save a corresponding decision preference map according to the decision preference content of the target task;

[0227] A decision unit generation unit, used to generate an intelligent aggregation decision execution unit for a target task;

[0228] The execution unit is used to make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute the corresponding target decision tasks according to the decision results.

[0229] A decision preference data processing device provided in an embodiment of the present invention can execute a decision preference data processing method provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method. It is worth noting that in the above-mentioned embodiment of the decision preference data processing device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0230] It can be seen that the data processing device described in the embodiment of the present application can improve the efficiency of executing the target task.

[0231] In a possible example, in the detection of the preset content, the comparison detection unit is specifically used to:

[0232] Get the decision preference presets for the target task.

[0233] In a possible example, in the aspect of executing the target task to synchronize the target operation in the server, the execution unit is specifically configured to:

[0234] Add the target task to a waiting queue;

[0235] Taking any task in the waiting queue at every preset time interval and loading the any task into the execution queue;

[0236] When the target task is obtained, the target task is executed to synchronize the target operation in the server.

[0237] In a possible example, the execution unit is further specifically configured to:

[0238] When the target task fails to be executed, the target task is transferred to a failure queue; or, when the target task is successfully executed, the target operation in the server is synchronized.

[0239] In one possible example, after adding the target task to the waiting queue corresponding to the target operation, and before taking any task in the waiting queue at every preset time interval, the execution unit is also specifically used to: extract network status parameters; evaluate network efficiency; and determine the preset time interval corresponding to the reference network rate according to the mapping relationship between the preset network rate and the time interval.

[0240] Embodiment 3

[0241] refer to Figure 4 , Figure 4 : is a structural diagram of a decision preference data processing device of this embodiment. The decision preference data processing device 20 of this embodiment includes a processor 21, a memory 22, and a computer program stored in the memory 22 and executable on the processor 21. When the processor 21 executes the computer program, the steps in the above method embodiment are implemented. Alternatively, when the processor 21 executes the computer program, the functions of each module / unit in the above device embodiments are implemented.

[0242] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the decision preference data processing device 20. For example, the computer program may be divided into the modules in Embodiment 2. For the specific functions of each module, please refer to the working process of the device described in the above embodiment, which will not be repeated here.

[0243] The decision preference data processing device 20 may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will appreciate that the schematic diagram is merely an example of the decision preference data processing device 20 and does not constitute a limitation on the decision preference data processing device 20. The decision preference data processing device 20 may include more or fewer components than shown in the diagram, or may combine certain components, or different components. For example, the decision preference data processing device 20 may also include input and output devices, network access devices, buses, and the like.

[0244] The processor 21 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor 21 is the control center of the decision preference data processing device 20, and uses various interfaces and lines to connect various parts of the entire decision preference data processing device 20.

[0245] The memory 22 can be used to store the computer program and / or module. The processor 21 realizes various functions of the decision preference data processing device 20 by running or executing the computer program and / or module stored in the memory 22 and calling the data stored in the memory 22. The memory 22 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory 22 can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0246] Wherein, if the module / unit integrated in the decision preference data processing device 20 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor 21, the steps of the above-mentioned method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0247] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0248] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 Processes Multiple processes and / or boxes Figure 1 A device that has the functions specified in one or more boxes.

[0249] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0250] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0251] The parts of the present invention that are not described in detail are prior art. It is obvious to those skilled in the art that the present invention is not limited to the details of the above-mentioned exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention; therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive, and it is intended that all changes that fall within the meaning and scope of equivalent elements are included in the present invention.

Claims

1. A decision preference data processing method, characterized in that: The following steps are involved: S1. Compare the detection target task with the decision preference preset content template to find the decision preference preset content of the target task; The decision preference preset content template includes a plurality of different types of preset decision preference content; The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags; Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task; S2, displaying the decision preference preset map corresponding to the decision preference preset content of the target task; S3, changing the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map; S4. Generate and save the corresponding decision preference map according to the decision preference content of the target task; S5, generating an intelligent aggregation decision execution unit for the target task; Step S5 is specifically as follows: To target tasks Aggregate the output corresponding utility, and then combine it with the judgment function Generate an intelligent aggregation decision execution unit for the target task; the aggregation function is as follows: ， in, for subtasks or attributes; It is for indivual Aggregation functions for scene usage; It is for indivual The weight of the scene; Indicates all The judgment function Used for different purposes Perform different target operations; S6. Make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute corresponding target decision tasks based on the decision results.

2. The method according to claim 1, characterized in that: In step S1, the matching degree between the target task and the preset content of the decision preference is calculated according to the label and the label weight, which specifically includes: ， in, Preset weights of content labels for decision preferences; Preset matching values ​​of content labels for target tasks and decision preferences; Indicates the sequence number of the target task; A sequence number for presetting content labels for decision preferences; is the number of labels.

3. The method according to claim 1, characterized in that Step S1 also includes: performing semantic understanding and matching calculation on the target task description and labels through a pre-trained language model.

4. The method according to claim 1, characterized in that: Step S3 specifically includes: obtaining touch parameters of the decision preference preset map display area, and adjusting the preference parameters by touch.

5. The method according to claim 4, characterized in that There is a mapping relationship between the touch parameters and the preference parameters. When the touch parameters change, the decision preference preset map changes synchronously with the preference parameters.

6. The method according to claim 1, characterized in that Step S6 also includes: when the target task fails to be executed, transferring the target task to a failure queue; when the target task is successfully executed, synchronizing the decision preference content of the target task to a decision preference preset content template.

7. The method according to claim 1, characterized in that Step S4 also includes: evaluating the network status when network transmission operations are involved: calculating the comprehensive network efficiency score using a weighted summation formula, grading the network according to the comprehensive network efficiency score, and taking different operations for networks of different levels; The overall network efficiency score The formula is: ， in, As an indicator, is the actual value of the indicator; is the weight of the indicator; is the standard deviation of the indicator; is the nonlinear index; is the dynamic adjustment factor; is the maximum theoretical score based on all possible network states; when When the network status is excellent, data transmission and high bandwidth demand tasks are fully performed; when and When , the network status is good network, and the balanced data is synchronized; when and When the network status is a poor network, only the key data is updated and the transmission of non-key data is restricted; the key data includes: decision preference data adjusted by the user, task-related data input by the user; the non-key data includes: calculation process data, statistical data; when , the network status is poor, at this time, data synchronization is restricted, local operations are recorded first, and the network status is poor.

8. A decision preference data processing device, characterized in that: The following units are included: A comparison and detection unit, used for including a plurality of different types of preset decision preference contents in the decision preference preset content template; The decision preference preset content includes a name, a tag set and a tag weight set; the tags are multiple and different; the sum of all elements in the tag weight set is 1; the number of the tag weights is consistent with the number of the tags; Calculating the matching degree between the target task and the decision preference preset content according to the label and the label weight, sorting according to the matching degree, and selecting the decision preference preset content with the highest matching degree as the decision preference preset content of the target task; A display unit, used to display a decision preference preset map corresponding to the decision preference preset content of the target task; An adjustment unit, used to change the preference parameters of the decision preference preset content of the target task by adjusting the decision preference preset map; A preference map generation unit, used to generate and save a corresponding decision preference map according to the decision preference content of the target task; A decision unit generation unit, used to generate an intelligent aggregation decision execution unit for a target task; Through the aggregation function To target tasks Aggregate the output corresponding utility, and then combine it with the judgment function Generate an intelligent aggregation decision execution unit for the target task; the aggregation function is as follows: ， in, for subtasks or attributes; It is for indivual Aggregation functions for scene usage; It is for indivual The weight of the scene; Indicates all The judgment function Used for different purposes Perform different target operations; The execution unit is used to make decisions and judgments on tasks that meet the characteristics of the target tasks through intelligent aggregation decision units and give decision results, and execute the corresponding target decision tasks according to the decision results.

9. An electronic device, characterized in that: The method comprises a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for executing the steps in the method according to any one of claims 1 to 7.

Images