CN116955292A - EAST algorithm-based digital chemical engineering bill generation method and system - Google Patents

Abstract

The invention relates to the technical field of digital systems, and discloses a digital chemical engineering bill generation method and a digital chemical engineering bill generation system based on an EAST algorithm. According to the method and the system for generating the digital chemical engineering bill based on the EAST algorithm, the data are subjected to the weight dividing processing by inserting the weight dividing processing algorithm into the system, the influence degree of different data in different units is highlighted, and corresponding data can be found quickly by inputting the key parts of the data.

Description

Digital work order generation method and system based on EAST algorithm

Technical field

The invention relates to the field of digital systems, specifically a digital work order generation method and system based on the EAST algorithm.

Background technique

Digitization is to transform a lot of complex and changeable information into measurable numbers and data, and then use these numbers and data to establish appropriate digital models, transform them into a series of binary codes, introduce them into the computer for unified processing, this is The basic process of digitization, which converts any continuously changing input such as the lines of a picture or a sound signal into a string of separated units, represented by 0s and 1s in computers.

Today's work orders are too cumbersome to search. You need to enter keywords one by one to search, and there is a lot of data to be found, which greatly affects the search speed and efficiency. For this reason, there is an urgent need for a digital work order generation based on the EAST algorithm. Methods and systems.

Contents of the invention

In order to achieve the above objectives, the present invention provides the following technical solution: a digital work order generation method and system based on the EAST algorithm, including the following steps:

S1: Organize the data, the staff will classify and organize the data from different units;

S2: Input information, the staff will enter the compiled information into the system;

S3: Decentralized processing, the system decentralizes the processing of data;

S4: Save the data. The system will generate different digital units from the processed data and save them in classified intervals;

S5: Mobilize data, open intervals at different locations and enter key data for search.

Preferably, in step s4, the system generates different digital units from the processed data, and encrypts and hides the folders.

Preferably, in step s5, during the process of transferring data, the folder is opened, the folder is called out, the folder is decrypted, and the text is processed by inputting key text.

Preferably, in step s3, the decentralized processing algorithm is as follows:

K＝(k1*a1+k2*a2+k3*a3+....+kn*an)/(k1+k2+k3+...+kn));

X＝(X1*a1+X2*a2+X3*a3+....+Xn*an)/(X1+X2+X3+...+Xn));

The coefficients (k1, k2, k3, ....kn) and (X1+X2+X3+...+Xn) are weights, indicating the proportion of the data behind the coefficients in the entire statistical data, and also indicating that this data influence on statistical results.

Insert a decentralized processing algorithm inside the system to decentralize data and highlight the impact of different data on different units. By inputting the key parts of the data, the corresponding information can be quickly found.

Preferably, the file is opened in text mode for normal writing, and the file is opened in binary mode for binary mode writing of encrypted content. The encryption equation is a*x^2+b, where a and b are short integers and are input by the user. value.

Preferably, in step s1, the staff organizes the data of each unit, places the data of different units in different locations in the folder, and marks the corresponding locations.

Preferably, in step s2, the staff labels the corresponding unit in a single folder, and then inputs the categories of each data in the folder, so that it can be found quickly during the search process.

A system for generating digital work orders based on the EAST algorithm, including a classification file module, an algorithm system module and a search module, and virtual digital human technology. The algorithm system module is connected to the internal network of the search module, and the search module includes a display module. , a cylinder is fixedly connected to the back of the display module, a telescopic cylinder is fixedly connected to the inner bottom wall of the cylinder, a support frame is movably connected to the top of the telescopic cylinder, a shaft is movably connected to the inside of the cylinder, and the support A support plate is movably connected to the side of the frame away from the telescopic cylinder.

A system for generating digital work orders based on the EAST algorithm. There is a hole slot inside the cylinder, the support frame is located inside the hole slot, the shaft is located inside the hole slot, and the outside of the shaft rod is in contact with the support frame. The internal movable connection is provided with four supporting frames, and the four supporting frames are located outside the cylinder and distributed in a circumferential array.

A support plate is provided on the other side of the support frame. The support frame is movably connected to the telescopic cylinder. The support frame is movably connected to the inside of the cylinder. The support frame is movably connected to the support plate. When the telescopic cylinder is extended, the support frame is located on the support plate. One side moves downward, and the two ends of the support frame move toward each other. The movement of the telescopic cylinder controls the flipping movement of the support rod and the support plate. When in use, the telescopic cylinder is extended upward, and the support plate can be well positioned. On the desktop, support the display module. On the contrary, shorten it when it needs to be transported, and put away the support plate and support frame for easy storage and storage.

A system for generating digital work orders based on the EAST algorithm. The virtual digital human technology includes an environmental resource module, an appearance resource module, a voice function module, an intelligent engine module and an intelligent switching module;

The environmental resource module, large screens, mobile phones, VR and other display devices are only used as media to carry digital intelligence. We hope that the interface will be "invisible" in specific scenes, so that digital intelligence can truly stand in the scene and gain user affinity and trust. Focus on face-to-face communication with her without feeling disobedient or intrusive. Regarding the design of the background of the digital sapien, the environmental resources simulate real offline customer service scenes, training scenes, festival scenes, and even handle details such as gestures and gestures;

The appearance resource module is to construct the face shape, facial features, hairstyle, makeup and clothing of a sapient. For a business sapient, it is necessary to think about what kind of face, facial features and clothing he needs to have from the aspects of personality traits and professional traits. Factors, based on different scenarios and corresponding service customers, digital smart people need to show smart and professional professional characteristics, and at the same time have a friendly personality. Referring to relevant facial psychology research, the appearance of thick eyebrows, big eyes, and a slightly pointed face will give The impression of a smart person, and an appearance with raised corners of the mouth, slightly curved eyes, and less prominent edges and corners of the face will be more friendly and lovable. We also try to be close to these characteristics when choosing appropriate adjustments to the image and clothing;

The speech function module has basic functions such as speech recognition, sound synthesis, intonation and other basic settings. It can restore the rhythm, laughter, paralanguage, etc. of real people in a fine-grained manner. It has such a strong sense of object that audiometers can no longer distinguish between synthesized speech and real people. Recording allows digital smart people to have their own basic vocal abilities so that they can cooperate with the back-end intelligent brain to realize intelligent questioning. The core of its call lies in AI, including speech recognition, semantic understanding, etc. Powerful AI interaction capabilities are the key to intelligent voice products. Core competitiveness; and the accumulation of large amounts of data can effectively improve model performance and enhance the interactive experience of virtual humans;

The intelligent engine module reflects the content and emotions in the communication of digital smart people, that is, it constructs the selection and expression of digital smart people's emotional actions and faster and more accurate communication content. Because business-type digital smart people mainly carry efficiency-oriented In terms of tasks, we have weakened the expression of the personality traits of Homo sapiens in the selection of emotions, words, deeds, and actions, and are more biased towards the expression of professional traits, allowing users to experience services effectively and pleasantly in real online and offline scenarios;

The intelligent interaction module recognizes the user's intention through intelligent technologies such as speech semantic recognition, and determines the subsequent voice and actions of the digital sapien based on the user's current intention, driving the character to start the next round of interaction. In terms of speech semantic understanding, it has Including speech recognition (ASR), semantic understanding (NLP/knowledge graph), speech synthesis (TTS) and many other technical capabilities, based on computer deep learning technology, allowing digital intelligence to continuously learn new corpora and data. After training and After receiving feedback from real-life applications, it continuously iterates and extends its capabilities and interaction levels. Through deep learning, digital intelligence not only deepens its professionalism, but also learns cross-business knowledge and supports cross-business services and communication.

Compared with the existing technology, the present invention provides a digital work order generation method and system based on the EAST algorithm, which has the following beneficial effects:

1. The digital work order generation method and system based on the EAST algorithm decentralizes the data by inserting a decentralized processing algorithm into the system, highlighting the impact of different data on different units. By inputting key parts of the data, Corresponding information can be found quickly.

2. The digital work order generation method and system based on the EAST algorithm can protect the data of the file through encryption on different folders and the binary encryption method to prevent the loss of data in the file.

3. The digital work order generation method and system based on the EAST algorithm sets a cylinder on the back of the display module, a telescopic cylinder inside the cylinder, a support frame on the top of the telescopic cylinder, and a support frame on the top of the support frame. A support plate is provided on the other side. The support frame is movably connected to the telescopic cylinder. The support frame is movably connected to the inside of the cylinder. The support frame is movably connected to the support plate. When the telescopic cylinder is extended, the support frame is located on one side of the support plate. downward movement, the two ends of the support frame move toward each other, and the movement of the telescopic cylinder controls the flipping movement of the support rod and the support plate. When in use, the telescopic cylinder is extended upward, and the support plate can be well positioned on the table. The display module is supported. On the contrary, when it needs to be transported, it should be shortened and the support plate and support frame should be put away for convenience.

Description of the drawings

Figure 1 is a flow chart of operating steps of the present invention;

Figure 2 is a schematic diagram of the operating steps of the present invention;

Figure 3 is a flow chart of the file encryption operation of the present invention;

Figure 4 is a schematic structural diagram of the back of the display module of the present invention;

Figure 5 is a schematic diagram of the internal structure of the cylinder of the present invention;

Figure 6 is a schematic diagram of the virtual digital human technology of the present invention.

Among them: 1. Display module; 2. Cylinder; 3. Telescopic cylinder; 4. Support frame; 5. Shaft; 6. Support plate.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

Embodiment 1

An embodiment of a digital work order generation method based on the EAST algorithm

Please refer to Figure 1-6, a digital work order generation method and system based on the EAST algorithm, including the following steps:

S1: Organize the data, the staff will classify and organize the data from different units;

S2: Input information, the staff will enter the compiled information into the system;

S3: Decentralized processing, the system decentralizes the processing of data;

S4: Save the data. The system will generate different digital units from the processed data and save them in classified intervals;

S5: Mobilize data, open intervals at different locations and enter key data for search.

In the s4 step, the system generates different digital units from the processed data, and encrypts and hides the folders.

In the s5 step, during the process of transferring data, the folder is opened to call out the folder, decrypted, and the text is processed by inputting key text.

In step s3, the decentralized processing algorithm is as follows:

K＝(k1*a1+k2*a2+k3*a3+....+kn*an)/(k1+k2+k3+...+kn));

X＝(X1*a1+X2*a2+X3*a3+....+Xn*an)/(X1+X2+X3+...+Xn));

The coefficients (k1, k2, k3, ....kn) and (X1+X2+X3+...+Xn) are weights, indicating the proportion of the data behind the coefficients in the entire statistical data, and also indicating that this data influence on statistical results.

In step s1, the staff organizes the data of each unit, places the data of different units in different locations in the folder, and marks the corresponding locations.

In the s2 step, the staff labels the corresponding units in a single folder, and then inputs the categories of each data in the folder, which can be found quickly during the search process.

Through the above technical solution, the decentralized processing algorithm is inserted into the system to decentralize the data and highlight the influence of different data on different units. By inputting the key parts of the data, the corresponding information can be quickly found.

Embodiment 2

An embodiment of encryption for digital work order generation method based on EAST algorithm

Open the file in text mode for normal writing, and open the file in binary mode for binary mode writing of encrypted content. The encryption equation is a*x^2+b, a and b are short integers, and the values are input by the user.

Through the above technical solution, by encrypting on different folders, the data of the file can be well protected through the binary encryption method to prevent the loss of data in the file. If the data written in text mode is read for the first time , after reading, before writing the encrypted content in binary mode, be sure to use the rewind() function, rewind(fp), to ensure that the encrypted content can be written only after the above operations.

Embodiment 3

An embodiment of a system for digital work order generation method based on EAST algorithm

A digital work order generation system based on the EAST algorithm, including a classification file module, an algorithm system module and a search module, and virtual digital human technology. The algorithm system module is connected to the internal network of the search module. The search module includes a display module 1, and the back of the display module 1 There is a cylinder 2 fixedly connected, a telescopic cylinder 3 is fixedly connected to the inner bottom wall of the cylinder 2, a support frame 4 is movably connected to the top of the telescopic cylinder 3, a shaft 5 is movably connected inside the cylinder 2, and the support frame 4 is away from the telescopic cylinder 3. A support plate 6 is movably connected to the side.

There is a hole slot inside the cylinder 2, the support frame 4 is located inside the hole slot, the shaft rod 5 is located inside the hole slot, the outside of the shaft rod 5 is movably connected to the inside of the support frame 4, there are four support frames 4, four support frames 4 They are located outside the cylinder 2 and distributed in a circular array.

Through the above technical solution, a cylinder 2 is provided on the back of the display module 1, a telescopic cylinder 3 is provided inside the cylinder 2, a support frame 4 is provided on the top of the telescopic cylinder 3, and on the other side of the support frame 4 A support plate 6 is provided. The support frame 4 is movably connected to the telescopic cylinder 3. The support frame 4 is movably connected to the inside of the cylinder 2. The support frame 4 is movably connected to the support plate 6. When the telescopic cylinder 3 is extended, the support frame 4 is located at One side of the support plate 6 moves downward, and the two ends of the support frame 4 move toward each other. The movement of the telescopic cylinder 3 controls the flipping movement of the support rod 4 and the support plate 6. When in use, the telescopic cylinder 3 is extended upward. , the support plate 6 can be well positioned on the desktop to support the display module 1. On the contrary, when it needs to be transported, it can be shortened and the support plate and the support frame can be put away for convenience.

Embodiment 4

An embodiment of a system for digital work order generation method based on EAST algorithm

The virtual digital human technology includes an environmental resource module, an appearance resource module, a voice function module, an intelligent engine module and an intelligent switching module;

The environmental resource module, large screens, mobile phones, VR and other display devices are only used as media to carry digital intelligence. We hope that the interface will be "invisible" in specific scenes, so that digital intelligence can truly stand in the scene and gain user affinity and trust. Focus on face-to-face communication with her without feeling disobedient or intrusive. Regarding the design of the background of the digital sapien, the environmental resources simulate real offline customer service scenes, training scenes, festival scenes, and even handle details such as gestures and gestures;

The appearance resource module is to construct the face shape, facial features, hairstyle, makeup and clothing of a sapient. For a business sapient, it is necessary to think about what kind of face, facial features and clothing he needs to have from the aspects of personality traits and professional traits. Factors, based on different scenarios and corresponding service customers, digital smart people need to show smart and professional professional characteristics, and at the same time have a friendly personality. Referring to relevant facial psychology research, the appearance of thick eyebrows, big eyes, and a slightly pointed face will give The impression of a smart person, and an appearance with raised corners of the mouth, slightly curved eyes, and less prominent edges and corners of the face will be more friendly and lovable. We also try to be close to these characteristics when choosing appropriate adjustments to the image and clothing;

The speech function module has basic functions such as speech recognition, sound synthesis, intonation and other basic settings. It can restore the rhythm, laughter, paralanguage, etc. of real people in a fine-grained manner. It has such a strong sense of object that audiometers can no longer distinguish between synthesized speech and real people. Recording allows digital smart people to have their own basic vocal abilities so that they can cooperate with the back-end intelligent brain to realize intelligent questioning. The core of its call lies in AI, including speech recognition, semantic understanding, etc. Powerful AI interaction capabilities are the key to intelligent voice products. Core competitiveness; and the accumulation of large amounts of data can effectively improve model performance and enhance the interactive experience of virtual humans;

The intelligent engine module reflects the content and emotions in the communication of digital smart people, that is, it constructs the selection and expression of digital smart people's emotional actions and faster and more accurate communication content. Because business-type digital smart people mainly carry efficiency-oriented In terms of tasks, we have weakened the expression of the personality traits of Homo sapiens in the selection of emotions, words, deeds, and actions, and are more biased towards the expression of professional traits, allowing users to experience services effectively and pleasantly in real online and offline scenarios;

The intelligent interaction module recognizes the user's intention through intelligent technologies such as speech semantic recognition, and determines the subsequent voice and actions of the digital sapien based on the user's current intention, driving the character to start the next round of interaction. In terms of speech semantic understanding, it has Including speech recognition (ASR), semantic understanding (NLP/knowledge graph), speech synthesis (TTS) and many other technical capabilities, based on computer deep learning technology, allowing digital intelligence to continuously learn new corpora and data. After training and After receiving feedback from real-life applications, it continuously iterates and extends its capabilities and interaction levels. Through deep learning, digital intelligence not only deepens its professionalism, but also learns cross-business knowledge and supports cross-business services and communication.

Through the above technical solutions and based on this demand, a multi-modal appearance resource recommendation algorithm based on multi-modal feature interaction is designed. This algorithm uses the Bert natural language processing model to vectorize the personality traits and professional traits of digital people and encode them in high dimensions. Two qualities that characterize the virtual digital human in the latent feature space; at the same time, the visual transformer model is used to vectorize the appearance resources of the digital human, and the encoding is also embedded in the high-dimensional latent feature space. In the high-dimensional latent feature space, use The cross-attention mechanism performs feature interaction on all codes embedded in this space to enhance the interconnection between virtual human traits and appearance resources; finally, the K nearest neighbor algorithm is used to perform similarity between trait vectorization coding and appearance resource vectorization coding. degree calculation, calculate and match the N appearance resources with the highest similarity, thereby recommending appropriate appearance resources for each digital person. The algorithm is a deep learning algorithm. Adversarial learning is used for training. The loss function is Tripletloss. The difficult sample mining strategy is used during the model training process to enhance the generalization and convergence acceleration of the model;

Based on this demand, an algorithm for exaggerating facial emotional expressions based on collaborative analysis is proposed. The virtual human face image is obtained through visual capture, and the FLAME model is used to unsupervisedly train facial expression blendshape estimation. The number of blendshapes is 300. After obtaining the Blendshape, use the self-attention mechanism (Self-attention) to actively learn to predict key expression base coefficients, use the DBSCAN algorithm to complete key expression base clustering, and amplify the expression coefficients of each cluster according to its criticality to complete expression exaggeration. Processed to highlight the key emotions of Homo sapiens. The algorithm is trained in an unsupervised manner throughout the entire process, and the supervision signals are 2D key points on the face;

In the environmental resources module

It can realize scene configuration according to different application scenarios, and the image, camera position, sound and other content of Homo sapiens can be set in the scene;

It can realize the setting of horizontal screen and vertical screen, and supports custom selection of resolution and bit rate;

It is possible to upload local pictures as the background pictures of virtual Homo sapiens, and supports the selection of multiple background pictures. Jpg/png format is required, and there is no limit on image size.

Appearance resource module

Virtual human image assets mainly include clothing, hairstyle, shoes, etc. Homo sapiens clothing assets can be produced based on clothing pictures provided by customers to restore the appearance of the original brand of clothing to the greatest extent and present the data. Clothing assets support material/color replaceable status;

The movements of virtual digital homosapiens are usually data collected through motion capture equipment, and then animators repair the collected data frame by frame. The repaired movements can be deployed in the product and called on demand during use;

Use matrix 3D scanning as the data basis to produce image resources, match the body shape/grid distribution density of virtual Homo sapiens, topology the high-model data, reduce the number of faces of the actual image resource model while retaining the accurate structure and structure of the model to the maximum extent level of sophistication;

Data is collected through motion capture equipment, and then animators perform frame-by-frame restoration on the collected data. During data collection, professional motion capture solutions are used, including light capture, habit capture, etc.;

Voice resource module

It can collect directional dry sound data based on the timbre of a specified speaker. The dedicated pronunciator goes to a professional recording studio to collect dry sound data, and the algorithm engineer conducts special training on the collected dry sound data to generate an enterprise-level customized sound library. Virtual Digital Homo TTS adopts cross-decoupling technology, fully decoupling style and timbre, and uses cross-language transfer technology to support Chinese and English bilingual synthesis. The sound quality has the characteristics of high restoration, high definition, and high stability synthesis;

It is possible to customize the speed, pitch, volume and other parameters of the voice of the Homo sapiens. For situations where there are multiple speakers, the voice can be selected as needed.

Intelligent engine module

The virtual digital intelligence population-driven algorithm is the core indicator for evaluating the simulation degree of digital intelligence. Virtual digital humans need to provide personalized interactions and services based on intelligent perception of drivers. In terms of expression and movement control, based on a lightweight AI model and a variety of preset smiles, eyebrow movements and preset body movements, it can achieve accurate expression of body movements, facial expressions and word-for-word mouth shapes;

The attribute management function allows the definition and editing of synonyms, corpora, types, units, terms, introductions, etc. for attributes in the knowledge inference engine. It also allows supplementary editing for situations where different entities have different types, units, and speech techniques when matching common attributes;

The multi-turn dialogue engine provides some existing scenarios for use. You can select an existing scene template and modify it on the basis of it, or you can directly design a new scene.

Configuration of basic elements for conducting multiple rounds of dialogue;

You can choose to import locally prepared scene scripts and make modifications and adjustments on this basis;

Help identify user emotions by creating emotion models. The built-in emotion model is reusable;

You can query based on time, module, search, score, auxiliary error correction, customer dialogue word count, whether to ignore, and status.

Intelligent interactive module

Answer category statistics provide insights into trends in business conversations as well as non-business conversations. Business conversations refer to conversations provided by the two modules of common standard questions and task engines; chat conversations refer to small talk; other categories refer to unknown replies, robot skills, sensitive word replies, etc.;

Hot questions display the 100 most frequently asked standard questions in the selected time period (today/yesterday/last 7 days/last 30 days) to help discover the most common business problems;

Provides editable and customizable chat corpus to customize the robot's chat questions and answers.

Environmental Resources Module

Through the virtual background environment, natural and real dialogue scenes are created, allowing users to experience services effectively and pleasantly in real offline scenes. Entering the situational communication stage, the above-mentioned State Grid elements are integrated and interpreted, and users can clearly establish the perception of "intelligent, consultable, trustworthy, and usable". At this time, combined with the internal factors of user consultation business needs, the experience can be improved impulse.

Appearance resource module

Construct appearance resources such as facial features, movements, stance, hairstyle, makeup and clothing of Homo sapiens. Show smart and professional professional characteristics, but also have a friendly personality. Referring to relevant facial psychology research, an appearance with thick eyebrows, big eyes, and a slightly pointed face will give people an impression of intelligence, while an appearance with raised corners of the mouth, slightly curved eyes, and less prominent face edges will be more friendly and lovable. When choosing a model, We also try to stay as close to these features as possible.

Intelligent engine module

Through the digital intelligent human driven engine, the action engine, mouth shape driver, question and answer engine, and expression driver are run in conjunction, truly achieving clear hearing, accurate translation, fast output, good adaptability to online and offline environments, and natural and efficient user interaction. .

Intelligent interactive module

At the level of intelligent interaction, before using NLP to pre-train a large model, the semantic understanding ability requires a large amount of data that matches the application scenario to train the model. If you want to copy the model capabilities to different scenarios on a large scale, you will face considerable data cost pressure. NLP The use of pre-trained large models significantly reduces the data requirements of the semantic understanding model in new business cold start scenarios. There are more intelligent interaction scenarios, such as single-round dialogues, multi-round dialogues, and task-based dialogues, which can cover business guidance and questions. Consulting, business handling, etc.

Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art will understand that various changes, modifications, and substitutions can be made to these embodiments without departing from the principles and spirit of the invention. and modifications, the scope of the invention is defined by the appended claims and their equivalents.

Claims (10)

1. A digital chemical engineering bill generation method based on EAST algorithm is characterized in that: the method comprises the following steps:

s1: sorting the data, wherein staff sorts the data of different units for sorting;

s2: inputting data, wherein a worker inputs the well-arranged data into a system;

s3: the system performs the weight dividing treatment on the data;

s4: storing data, wherein the system generates different digital units from the processed data, and stores the classified intervals;

s5: and mobilizing the data, opening the intervals at different positions, inputting key data, and searching.

2. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: in step s4, the system generates different digitizing units for the processed data, and encrypts and hides the file.

3. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: in the step s5, the folder is opened to call out the folder in the process of mobilizing the data, and the folder is decrypted, and characters are processed by inputting key characters.

4. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: in step s3, the split weight processing algorithm is as follows:

K＝(k1*a1+k2*a2+k3*a3+....+kn*an)/(k1+k2+k3+...+kn))；

X＝(X1*a1+X2*a2+X3*a3+....+Xn*an)/(X1+X2+X3+...+Xn))；

the coefficients (k 1, k2, k3,) kn and (x1+x2+x3+ & Xn) are weights indicating the specific gravity of the data following the coefficients in the whole statistical data and the degree of influence of this data on the statistical result.

5. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: the text mode opens the file, performs normal writing, and binary mode opens the file, performs binary mode writing of the encrypted content, and inputs the value by the user, wherein the encryption equation is a x 2+b, and a, b is a short integer.

6. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: in the step s1, the staff sorts the data of each unit, places the data of different units at different positions of the folder, and marks the corresponding positions.

7. The method for generating the digital chemical engineering bill based on the EAST algorithm as claimed in claim 1, wherein the method comprises the following steps: in the step s2, staff marks the corresponding units on a single folder, then inputs the classified data in the folder, and the classified data can be quickly found in the searching process.

8. A system of a digital chemical engineering bill generation method based on EAST algorithm is characterized in that: including categorised file module, algorithm system module and seek module, virtual digital man-machine technique, algorithm system module and seek module internal network connection, seek the module and include display module (1), display module (1) back fixedly connected with drum (2), wall fixedly connected with telescopic cylinder (3) in drum (2), telescopic cylinder (3) top swing joint has support frame (4), inside swing joint of drum (2) has axostylus axostyle (5), telescopic cylinder (3) one side swing joint has supporting disk (6) are kept away from to support frame (4).

9. The system of the digital chemical engineering bill generating method based on the EAST algorithm according to claim 8, wherein the system comprises the following components: the novel cylindrical drum is characterized in that a hole groove is formed in the cylinder (2), the supporting frame (4) is located inside the hole groove, the shaft rod (5) is located inside the hole groove, the outer side of the shaft rod (5) is movably connected with the inner side of the supporting frame (4), the four supporting frames (4) are arranged, and the four supporting frames (4) are located on the outer side of the cylinder (2) and are distributed in a circumferential array.

10. The system of the digital chemical engineering bill generating method based on the EAST algorithm according to claim 8, wherein the system comprises the following components: the virtual digital man-in-the-art technology comprises an environment resource module, an appearance resource module, a voice function module, an intelligent engine module and an intelligent exchange module;

the environment resource module, the large screen, the mobile phone, the VR and other display equipment are only used as media for bearing the digital wisdom, the interface is hoped to be hidden in a specific scene, the digital wisdom is enabled to truly stand in the scene, the affinity trust sense of the user is obtained, the user is focused on face-to-face communication, no offence and invasion are felt, and regarding the design of the digital wisdom background, the environment resource simulates the details of a real offline customer service scene, a practical training scene, a holiday scene, even gesture gestures and the like;

the appearance resource module is used for constructing facial features, hairstyles and dressing clothes of a plurality of wisdom people, for business-type wisdom people, the aspects of facial features, facial features and clothes elements which the wisdom people need to possess are needed to be considered from the aspects of personality traits and professional traits, based on different scenes and corresponding service clients, the wisdom people need to show professional features of intelligence and expertise, meanwhile, the appearance of the wisdom people has the character which is compatible with people, and the appearance of the wisdom eyes and the facial points gives the smart impression, the corners of the mouth are raised, the eyes are slightly bent, the appearance of the facies edges which are not prominent is more affinitive discussed, and the appearance of the wisdom people is as close to the features as possible on the selection of proper adjustment of the image clothes;

the voice function module has basic functions of voice recognition, voice synthesis, intonation and the like, fine granularity restores real human rhythm, laughing sound, auxiliary language and the like, has extremely strong object sense, and audiometrists can not distinguish synthesized voice from real human recording, so that a plurality of intelligent persons have own basic sounding capability so as to cooperate with a rear-end intelligent brain to realize intelligent inquiry, the calling core is AI (analog input) and comprises voice recognition, semantic understanding and the like, and the strong AI interaction capability is the core competitiveness of intelligent voice products; the accumulation of a large amount of data can effectively improve the model expression height and improve the interaction experience of virtual people;

the intelligent engine module reflects the content and emotion in the communication of the homo sapiens users, namely, the intelligent engine module constructs the type selection and expression of the homo sapiens emotion actions and the more rapid and accurate communication content, and as the business homo sapiens mainly bear the efficiency task, the emotion, the language and the action are selected to weaken the performance of the personality traits of the plurality of wisdom people, and the performance of the personality traits is favored, so that the user can effectively and pleasurably experience services in real online and offline scenes;

the intelligent interaction module recognizes the intention of a user through intelligent technologies such as voice semantic recognition and the like, determines subsequent voices and actions of the wisdom people according to the current intention of the user, drives the person to start the next round of interaction, has multiple technical capabilities including voice recognition (ASR), semantic understanding (NLP/knowledge graph), voice synthesis (TTS) and the like in terms of voice semantic understanding, enables the wisdom people to continuously learn new corpus and data based on a computer deep learning technology, and continuously iterates and extends the capabilities and interaction levels of the wisdom people after feedback of training and practical application, and can learn cross-service knowledge to a certain extent through deep learning to support cross-service and communication.