Full-process order life cycle management method, system, device and medium

Abstract

The invention relates to a full-process order life cycle management method, system and device and a medium, and the method comprises the following steps: obtaining user information and user demands, and building a work order; according to a preset rule, work order distribution is carried out, related information is supplemented, and a pre-sales scheme is generated; technical feasibility, resource requirements and risks of the pre-sales scheme are automatically assessed, and an assessment report is generated; project research and development are carried out based on the pre-sales scheme and the evaluation report, the research and development progress is updated in real time in the research and development process, and research and development results are uploaded after research and development are completed; the research and development results are tested based on the automatic test script, and a test report is automatically generated; generating a delivery list based on research and development results and test reports, and automatically sending the delivery list to the user; and obtaining questions fed back by the user, performing automatic classification according to the questions, and performing adjustment according to a classification result until the order is completed. Compared with the prior art, the device has the advantages of high automation degree and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of order lifecycle management, and in particular to a full-process order lifecycle management method, system, equipment and medium. Background Art

[0002] The existing intelligent equipment work order management system usually processes sales orders, R&D work orders, and after-sales work orders separately. The data interaction and process sharing between systems are poor, resulting in serious information island phenomenon, difficulty in real-time tracking of project progress, and slow response to customer needs. Specifically, there are the following defects:

[0003] ① The lack of effective data interaction mechanism between systems in various links leads to serious information island phenomenon and untimely and inaccurate information transmission;

[0004] ② The real-time and transparency of project progress is insufficient, making it difficult to achieve cross-departmental collaborative management;

[0005] ③ The response speed to customer needs is slow, making it difficult to meet the rapidly changing market demands;

[0006] ④Every process of order management requires manual intervention, and the degree of automation is not high. Summary of the invention

[0007] The purpose of the present invention is to overcome the problem that the existing work order system applicable to the smart device industry is relatively scattered and project management is prone to omissions, and to provide a full-process order life cycle management method, system, equipment and medium, which brings together the entire process of sales order creation-pre-sales plan-project evaluation-R&D order completion-after-sales follow-up, which is convenient for tracking the progress of customer demand processing, and is also conducive to the centralized management of customers by the project, connecting the three links of sales, R&D, and after-sales, enhancing the information connectivity and sharing between each other, realizing automatic management and real-time tracking of the entire process, improving the response speed to customer needs, reducing the phenomenon of information islands, and improving the transparency and collaborative efficiency of project management.

[0008] The purpose of the present invention can be achieved by the following technical solutions:

[0009] According to a first aspect of the present invention, a full-process order lifecycle management method is provided, the method comprising the following steps:

[0010] S1, obtain user information and user needs, and create a work order;

[0011] S2, assign work orders according to preset rules, supplement relevant information, generate pre-sales plans, and update the work order status to pending evaluation;

[0012] S3. Automatically evaluate the technical feasibility, resource requirements, and risks of the pre-sales solution, and generate an evaluation report. If the evaluation conclusion in the evaluation report is "passed", then execute step S4 and update the work order status to "under development". If the evaluation conclusion is "rejected", then generate the reason for rejection, and return to step S2 to modify the pre-sales solution and update the work order status to "pre-sales modification".

[0013] S4. Conduct project R & D based on the pre-sales solution and the evaluation report. During the R & D process, update the R & D progress in real-time, and upload the R & D results after R & D is completed, and update the work order status to "awaiting testing".

[0014] S5. Test the R & D results based on the automated test script, and automatically generate a test report. If the test passes, then upload the test report and update the work order status to "awaiting delivery". If the test fails, then return to step S4 to conduct R & D again.

[0015] S6. Generate a delivery list based on the R & D results and the test report, automatically send it to the user, and update the work order status to "delivered".

[0016] S7. Obtain the problems feedback by the user, and automatically classify them according to the problems. If the classification result is "problem type", then return to step S4 to conduct R & D again according to the feedback problems. If the classification result is "requirement type", then return to step S2 to generate a new pre-sales solution according to the feedback problems. If all problems have been solved or there are no problems, then update the work order status to "completed".

[0017] As an optimal technical solution, S1 includes the following steps:

[0018] Obtain user information and enter it into the work order. The user information includes the user name, contact information, and industry.

[0019] Obtain user requirements, and enter them into the work order in text form or by selecting through a template.

[0020] Determine the associated products or services according to the user requirements, and set the expected delivery time and priority, and enter them into the work order.

[0021] Generate a unique work order number according to the work order information.

[0022] As an optimal technical solution, S2 includes the following steps:

[0023] Allocate the work order according to the preset rules. The preset rules are to allocate according to the product type and / or priority.

[0024] Automatically fill in the user information in the pre-sales solution according to the information in the work order.

[0025] Configure the product parameters and quotes according to the user requirement information in the work order.

[0026] Obtain the uploaded requirement attachments, where the requirement attachments include technical documents and schematic diagrams;

[0027] Combine the user information, configured product parameters, quotations, and requirement attachments to generate a requirement solution.

[0028] As a preferred technical solution, the technical feasibility, resource requirements, and risks of the pre-sales solution are automatically evaluated using a multi-task neural network model, including the following steps:

[0029] Perform data cleaning and preprocessing on the data in the requirement document;

[0030] Perform feature encoding and standardization processing on the cleaned data to obtain data features;

[0031] Input the standardized data features into the trained multi-task neural network model, and output the evaluation results of technical feasibility, resource requirements, and risks; among them, the multi-task neural network model includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is the same as the number of data features. The output layer includes three branches, which respectively correspond to the evaluation results of technical feasibility, resource requirements, and risks. The technical feasibility branch uses one neuron, and maps the output to the interval [0,1] through an activation function, representing the probability of the project being technically feasible. The resource requirements branch uses one neuron to output the estimated value of the project's resource requirements. The risk branch uses multiple neurons, each neuron corresponding to a risk level, and outputs the probability corresponding to each risk level through an activation function.

[0032] As a preferred technical solution, when the probability of the project being technically feasible output by the technical feasibility branch is less than the preset threshold, or the estimated value of the resource requirements output by the resource requirements branch is greater than the preset threshold, or the highest risk level output by the risk branch is greater than the preset level, the evaluation result is unqualified.

[0033] As a preferred technical solution, the loss function of the multi-task neural network model is expressed as:

[0034] L = w1L1 + w2L2 + w3L3

[0035] Among them, L1 represents the technical feasibility loss, L2 represents the resource requirements loss, L3 represents the risk assessment loss, and w1, w2, and w3 are the weights corresponding to the losses.

[0036] The technical feasibility loss is expressed using the binary cross-entropy loss function:

[0037] L1 = -ylog(p) - (1 - y)log(1 - p)

[0038] Among them, y is the true label of technical feasibility, y = 0 represents infeasible, y = 1 represents feasible, and p is the probability of technical feasibility predicted by the model;

[0039] The loss of resource requirements is expressed using the mean squared error loss function:

[0040]

[0041] Among them, y i is the actual resource requirement, is the resource requirement predicted by the model, and n is the number of samples;

[0042] The risk assessment loss is expressed using the cross-entropy loss function:

[0043]

[0044] Among them, C is the number of risk levels, b i is the true risk level label, and p i is the probability of the i-th risk level predicted by the model.

[0045] As a preferred technical solution, the multi-task neural network model is trained using a dynamic weighting method, and the weights are dynamically adjusted according to the learning situation of each task in different training stages, specifically including the following steps:

[0046] Initialize weights: At the beginning of training, assign initial weights to the loss functions of each task

[0047] Weight adjustment during training: After each training cycle, calculate the loss change rate of each task

[0048]

[0049] Among them, represents the loss of the i-th task in the t-th training cycle, and the loss change rate of the first training cycle is initialized to a preset value;

[0050] Adjust the weights of each task based on the loss change rate:

[0051]

[0052] Among them, is the weight of the i-th task in the t-th training cycle, α is a hyperparameter used to control the amplitude of weight adjustment, and m is the number of tasks.

[0053] According to the second aspect of the present invention, there is provided a full-process order life cycle management system for implementing the method described above. The system includes:

[0054] An information acquisition and work order establishment module: acquires user information and user requirements and establishes a work order;

[0055] A pre-sales solution generation module: allocates the work order according to preset rules, supplements relevant information, generates a pre-sales solution, and updates the work order status to pending evaluation;

[0056] An evaluation module: automatically evaluates the technical feasibility, resource requirements, and risks of the pre-sales solution and generates an evaluation report. If the evaluation conclusion in the evaluation report is passed, the R & D progress synchronization module is called, and the work order status is updated to in R & D. If the evaluation conclusion is rejected, the reason for rejection is generated, and the pre-sales solution generation module is called to modify the pre-sales solution, and the work order status is updated to pre-sales modification;

[0057] An R & D progress synchronization module: conducts project R & D based on the pre-sales solution and the evaluation report, updates the R & D progress in real time during the R & D process, and uploads the R & D results after the R & D is completed, and updates the work order status to pending testing;

[0058] A testing module: tests the R & D results based on automated test scripts, automatically generates a test report. If the test passes, the test report is uploaded, and the work order status is updated to pending delivery. If the test fails, the R & D progress synchronization module is called to re-conduct R & D and update the progress;

[0059] A delivery module: generates a delivery list based on the R & D results and the test report, automatically sends it to the user, and updates the work order status to delivered;

[0060] A feedback adjustment module: acquires the problems feedback by the user and automatically classifies them according to the problems. If the classification result is a problem type, the R & D progress synchronization module is called to re-conduct R & D according to the feedback problems. If the classification result is a requirement type, the pre-sales solution generation module is called to re-generate the pre-sales solution according to the feedback problems; if all the problems have been solved or there are no problems, the work order status is updated to completed.

[0061] According to the third aspect of the present invention, there is provided an electronic device including a memory and a processor. A computer program is stored on the memory, and when the processor executes the program, the method described above is implemented.

[0062] According to the fourth aspect of the present invention, there is provided a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the method described above is implemented.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1. Full-process automated management: Through the automated adjustment and modification of work order status, the present invention realizes seamless connection and automated management in links such as sales, R & D, and after-sales.

[0065] 2. Real-time data interaction and sharing: Through the cross-departmental data interaction mechanism, real-time sharing and collaborative management of information are realized.

[0066] 3. Quick response to customer needs: Through the system's automated demand processing and work order transfer, the response speed to customer needs is improved.

[0067] 4. Real-time tracking and early warning of project progress: The present invention tracks the project progress in real time and timely feedbacks project information.

[0068] 5. The present invention can automatically evaluate projects without manual verification one by one, improving the degree of automation, and the evaluation results have higher reliability. Description of the Drawings

[0069] Figure 1 It is a flowchart of the method of the present invention. Detailed Embodiments

[0070] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0071] Embodiment 1

[0072] This embodiment provides a full-process order life cycle management method, as Figure 1 shown, the method includes the following steps:

[0073] S1. Obtain user information and user needs, and create a work order.

[0074] Specifically, S1 includes the following steps:

[0075] S11. Obtain user information and input it into the work order by the salesperson. In this embodiment, the obtained user information includes but is not limited to user name, contact information, and industry.

[0076] S12. Obtain user needs and input them into the work order in text form or by template selection.

[0077] S13. Determine the associated product or service according to the user needs, set it in the created work order through the drop-down menu or search function, and set the expected delivery time and priority (high / medium / low), and input it into the work order.

[0078] S14. After the work order is entered, a unique work order number (such as SL-2023-001) is automatically generated according to the work order information, and the user information is stored in the customer table of the database, and the work order information is stored in the work order table.

[0079] S2. Allocate the work order according to the preset rules, supplement the relevant information, generate a pre-sales plan, and update the work order status to pending evaluation.

[0080] Specifically, S2 includes the following steps:

[0081] S21. After generating the unique work order number, allocate the work order according to the preset rules. In this embodiment, the preset rules are to allocate according to the product type and / or priority; that is, allocate to the corresponding pre-sales personnel according to the product type, or, according to the product type and priority, consider the response timeliness of the pre-sales personnel for allocation. After the allocation is completed, an email / sms / system message is automatically sent to notify the pre-sales personnel.

[0082] S22. Automatically fill in the user information in the pre-sales plan according to the information in the work order.

[0083] S23. The pre-sales personnel contact the customer through the built-in communication tool in the system (such as the IM module) according to the user requirement information in the work order, supplement the requirement details, and configure the product parameters and quotes (support formula calculation).

[0084] S24. Obtain the uploaded requirement attachments, and the requirement attachments include technical documents and schematic diagrams.

[0085] S25. Combine the user information, configured product parameters, quotes and requirement attachments to generate a requirement plan. When the requirement plan is generated, the system automatically updates the work order status to pending evaluation, and notifies the project evaluation team by email / sms / system message.

[0086] S3. Automatically evaluate the technical feasibility, resource requirements and risks of the pre-sales plan, and generate an evaluation report. If the evaluation conclusion in the evaluation report is passed, execute step S4, and update the work order status to in R & D. If the evaluation conclusion is rejected, generate the rejection reason, and return to step S2 to modify the pre-sales plan, and update the work order status to pre-sales modification.

[0087] In this embodiment, the multi-task neural network model is used to automatically evaluate the technical feasibility, resource requirements and risks of the pre-sales plan, including the following steps:

[0088] S31. Perform data cleaning and preprocessing on the data in the requirement document, remove duplicate data, handle missing values (mean filling, median filling or filling according to business logic can be used) and outliers (such as identifying and handling through the IQR method).

[0089] S32. Perform feature encoding and normalization on the cleaned data to obtain data features. Among them, for categorical features, use one-hot encoding to convert them into numerical features. For ordinal categorical features (such as technical difficulty level), label encoding can be used.

[0090] S33. Input the normalized data features into the trained multi-task neural network model and output the evaluation results of technical feasibility, resource requirements, and risks.

[0091] The multi-task neural network model includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is the same as the number of data features. In this embodiment, three fully connected hidden layers are set, and the number of neurons in each hidden layer is 128, 64, and 32 respectively. After each hidden layer, the ReLU activation function is used to increase the non-linear expression ability of the model. The output layer includes three branches, corresponding to the evaluation results of technical feasibility, resource requirements, and risks respectively. The technical feasibility branch uses one neuron, and the output is mapped to the interval [0,1] through the sigmoid activation function, representing the probability of project technical feasibility. The resource requirements branch uses one neuron to output the estimated value of project resource requirements. The risk branch uses multiple neurons, each neuron corresponding to a risk level, and the probability corresponding to each risk level is output through the softmax activation function.

[0092] When the project technical feasibility probability output by the technical feasibility branch is less than the preset threshold, or the resource requirement estimated value output by the resource requirements branch is greater than the preset threshold, or the highest risk level output by the risk branch is greater than the preset level, the evaluation result is not passed.

[0093] The loss function of the multi-task neural network model is expressed as:

[0094] L = w1L1 + w2L2 + w3L3

[0095] Among them, L1 represents the technical feasibility loss, L2 represents the resource requirement loss, L3 represents the risk assessment loss, and w1, w2, and w3 are the weights corresponding to the losses.

[0096] The technical feasibility loss is expressed using the binary cross-entropy loss function:

[0097] L1 = -ylog(p) - (1 - y)log(1 - p)

[0098] Among them, y is the true label of technical feasibility, y = 0 represents infeasible, y = 1 represents feasible, and p is the technical feasibility probability predicted by the model.

[0099] The loss of resource requirements is represented by the mean squared error loss function:

[0100]

[0101] where y i is the actual resource requirement, is the resource requirement predicted by the model, and n is the number of samples;

[0102] The risk assessment loss is represented by the cross-entropy loss function:

[0103]

[0104] where C is the number of risk levels, b i is the true risk level label, and p i is the probability of the i-th risk level predicted by the model.

[0105] Traditional multi-task learning models usually simply sum the losses of each task with weights. In this embodiment, the multi-task neural network model is trained using a dynamic weighting method, which dynamically adjusts the weights according to the learning situation of each task at different training stages. At the beginning of training, for tasks that are easy to learn (such as technical feasibility classification), the weights can be appropriately reduced, so that the model can pay more attention to tasks that are difficult to learn (such as resource requirement prediction); as training progresses, the weights are dynamically adjusted according to the loss change of each task, so that the model can learn more effectively at different stages. Specifically, it includes the following steps:

[0106] 1. Initialize weights: At the start of training, assign initial weights to the loss functions of each task In this embodiment, they are set to 0.2, 0.6, and 0.2 respectively.

[0107] 2. Weight adjustment during training: After each training cycle, calculate the loss change rate of each task

[0108]

[0109] where represents the loss of the i-th task in the t-th training cycle, and the loss change rate of the first training cycle is initialized to a relatively large preset value (such as the initial average loss of all samples on this task);

[0110] Adjust the weights of each task based on the loss change rate:

[0111]

[0112] where is the weight of the i-th task in the t-th training cycle, α is a hyperparameter used to control the magnitude of weight adjustment, and m is the number of tasks.

[0113] The principle of the above weight adjustment formula is: if the loss change rate of a task is large (i.e., the learning effect of this task is poor), then the exponential term will be relatively large, so that the weight of this task in the next training cycle increases; conversely, if the loss change rate of a task is small (i.e., the learning effect of this task is good), then its weight will decrease relatively.

[0114] In this embodiment, the preprocessed data is divided into a training set, a validation set, and a test set according to the ratio of 70%, 15%, and 15%. The training set is used to train the model, and the parameters are updated using Stochastic Gradient Descent (SGD) or its variants (such as the Adam optimizer). After each training cycle, the validation set is used to evaluate the model performance, and the weights are dynamically adjusted according to the loss of each task on the validation set. To prevent overfitting, an early stopping strategy is adopted. If the loss on the validation set does not decrease for several consecutive cycles, the training is stopped.

[0115] The trained model is evaluated using the test set. For technical feasibility evaluation, metrics such as accuracy, precision, recall, and F1-score are used; for resource requirement prediction, metrics such as Mean Squared Error (MSE), Mean Absolute Error (MAE), etc. are used; for risk assessment, metrics such as accuracy and confusion matrix are used.

[0116] S4. Based on the pre-sales plan and the evaluation report, carry out project R & D. During the R & D process, the R & D progress is updated in real time, and after the R & D is completed, the R & D results are uploaded, and the work order status is updated to pending testing.

[0117] The evaluation team fills in the estimated R & D cycle according to the evaluation results, and the work order is assigned to the R & D team, and the work order status is updated to in R & D. The R & D team receives the work order, views the pre-sales plan and the evaluation report, creates R & D sub-tasks in the system (such as module development, test case design), updates the progress in real time (such as "50% in development"), and records problems. After the R & D is completed, the deliverables (code library link, firmware file, etc.) are uploaded, the work order status is updated to pending testing, and the test team is automatically notified. In one embodiment, integration with tools such as GitLab / Jira is supported to automatically synchronize the R & D progress.

[0118] S5. Based on the automated test script, test the R & D results, automatically generate a test report. If the test passes, the test report is uploaded, and the work order status is updated to pending delivery. If the test fails, return to step S4 to conduct R & D again.

[0119] The tester receives the work order, downloads the R & D deliverables, executes the automated test script (system-built test framework), automatically generates a test report template, and reduces manual entry.

[0120] S6. Generate a delivery list (including products, documents, training PPT, etc.) based on the R & D results and the test report, automatically send it to the user, and update the work order status to delivered.

[0121] After receiving the delivery list, the customer signs the acceptance form online (electronic signature function) and uploads the customer feedback (such as "the interface operation is complex").

[0122] S7. Obtain the problems feedback by the user, and automatically classify them according to the problems. If the classification result is a problem type (such as fault repair), then return to step S4 to re-develop according to the feedback problems. If the classification result is a requirement type (such as function upgrade), then return to step S2 to re-generate the pre-sales plan according to the feedback problems. If all the problems have been solved or there are no problems, then update the work order status to completed.

[0123] Embodiment 2

[0124] The above is the introduction of the method embodiment. The following further illustrates the solution of the present invention through a system embodiment.

[0125] A full-process order life cycle management system for implementing the method described in Embodiment 1, the system includes:

[0126] Information acquisition and work order establishment module: acquire user information and user requirements, and establish a work order;

[0127] Pre-sales plan generation module: allocate work orders according to preset rules, supplement relevant information, generate a pre-sales plan, and update the work order status to to be evaluated;

[0128] Evaluation module: automatically evaluate the technical feasibility, resource requirements and risks of the pre-sales plan, and generate an evaluation report. If the evaluation conclusion in the evaluation report is passed, then call the R & D progress synchronization module and update the work order status to in R & D. If the evaluation conclusion is rejected, then generate the rejection reason and call the pre-sales plan generation module to modify the pre-sales plan, and update the work order status to pre-sales modification;

[0129] R & D progress synchronization module: conduct project R & D based on the pre-sales plan and the evaluation report, update the R & D progress in real time during the R & D process, and upload the R & D results after the R & D is completed, and update the work order status to to be tested;

[0130] Test Module: Test the R & D achievements based on the automated test scripts, automatically generate test reports. If the test passes, upload the test report and update the work order status to pending delivery. If the test fails, call the R & D progress synchronization module to re - conduct R & D and update the progress;

[0131] Delivery Module: Generate a delivery list based on the R & D achievements and test reports, automatically send it to the user, and update the work order status to delivered;

[0132] Feedback and Adjustment Module: Obtain the problems feedback by the user, and automatically classify them according to the problems. If the classification result is a problem type, call the R & D progress synchronization module to re - conduct R & D according to the feedback problems. If the classification result is a requirement type, call the pre - sales solution generation module to regenerate the pre - sales solution according to the feedback problems. If all the problems have been solved or there are no problems, update the work order status to completed.

[0133] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0134] Embodiment 3

[0135] The electronic device of the present invention includes a central processing unit (CPU), which can execute various appropriate actions and processes according to the computer program instructions stored in the read - only memory (ROM) or the computer program instructions loaded from the storage unit into the random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM, and RAM are connected to each other through a bus. The input / output (I / O) interface is also connected to the bus.

[0136] Multiple components in the device are connected to the I / O interface, including: an input unit, such as a keyboard, a mouse, etc.; an output unit, such as various types of displays, speakers, etc.; a storage unit, such as a disk, an optical disc, etc.; and a communication unit, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit allows the device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0137] The processing unit executes the various methods and processes described above, such as methods S1 to S7. For example, in some embodiments, methods S1 to S7 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed onto the device via the ROM and / or the communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S7 described above may be executed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S7 by any other suitable means (e.g., by means of firmware).

[0138] The functions described above herein may be performed at least in part by one or more hardware logic components. By way of example and not limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system on a chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0139] The program code for implementing the methods of the present invention may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on the remote machine or server.

[0140] In the context of the present invention, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0141] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A full-process order lifecycle management method, characterized in that: The method comprises the following steps: S1, obtain user information and user needs, and create a work order; S2, assign work orders according to preset rules, supplement relevant information, generate pre-sales plans, and update the work order status to pending evaluation; S3, automatically evaluate the technical feasibility, resource requirements and risks of the pre-sales plan, and generate an evaluation report. If the evaluation conclusion in the evaluation report is passed, execute step S4 and update the work order status to under development. If the evaluation conclusion is rejected, generate the rejection reason, return to step S2 to modify the pre-sales plan, and update the work order status to pre-sales modification; S4, conduct project development based on pre-sales solutions and evaluation reports, update the development progress in real time during the development process, upload the development results after the development is completed, and update the work order status to pending testing; S5, testing the R&D results based on the automated test script, automatically generating a test report, uploading the test report if the test passes, and updating the work order status to pending delivery, if the test fails, returning to step S4 to conduct R&D again; S6, generates a delivery list based on R&D results and test reports, automatically sends it to the user, and updates the work order status to delivered; S7, obtain the problems reported by users and automatically classify them according to the problems. If the classification result is a problem type, return to step S4 and re-perform research and development based on the reported problems. If the classification result is a demand type, return to step S2 and regenerate the pre-sales plan based on the reported problems. If all problems have been solved or there are no problems, update the work order status to completed.

2. A full-process order lifecycle management method according to claim 1, characterized in that: The S1 comprises the following steps: Obtain user information and enter a work order, wherein the user information includes user name, contact information, and industry; Obtain user requirements and enter work orders in text form or through template selection; Determine the associated products or services based on user needs, set the estimated delivery time and priority, and enter the work order; Generate a unique work order number based on the work order information.

3. A full-process order lifecycle management method according to claim 1, characterized in that: The S2 comprises the following steps: Allocate work orders according to preset rules, where the preset rules are allocation according to product type and / or priority; Automatically fill in user information in the pre-sales plan based on the information in the work order; Configure product parameters and quotations based on user demand information in the work order; Obtain uploaded requirement attachments, which include technical documents and schematic diagrams; Generate a demand plan by combining user information, configured product parameters, quotations and demand attachments.

4. A full-process order lifecycle management method according to claim 3, characterized in that: The automatic evaluation of the technical feasibility, resource requirements and risks of the pre-sales solution is implemented using a multi-task neural network model, including the following steps: Perform data cleaning and preprocessing on the data in the requirement document; Perform feature encoding and standardization on the cleaned data to obtain data features; The standardized data features are input into the trained multi-task neural network model to output the evaluation results of technical feasibility, resource requirements and risks; wherein the multi-task neural network model includes an input layer, a hidden layer and an output layer, the number of neurons in the input layer is the same as the number of data features, the output layer includes three branches, corresponding to the evaluation results of technical feasibility, resource requirements and risks respectively, the technical feasibility branch uses one neuron, and the output is mapped to the [0,1] interval through an activation function, indicating the probability of technical feasibility of the project, the resource requirement branch uses one neuron, and outputs the estimated value of the resource requirement of the project, the risk branch uses multiple neurons, each neuron corresponds to a risk level, and the probabilities corresponding to each risk level are output respectively through the activation function.

5. A full-process order lifecycle management method according to claim 4, characterized in that: When the probability of technical feasibility of the project output by the technical feasibility branch is less than the preset threshold, or the estimated resource requirement value output by the resource requirement branch is greater than the preset threshold, or the risk level with the highest probability output by the risk branch is greater than the preset level, the evaluation result is failure.

6. A full-process order lifecycle management method according to claim 4, characterized in that: The loss function of the multi-task neural network model is expressed as: L=w1L1+w2L2+w3L3 Among them, L1 represents the loss of technical feasibility, L2 represents the loss of resource demand, L3 represents the loss of risk assessment, and w1, w2, and w3 are the weights of the corresponding losses. The technical feasibility loss is expressed using a binary cross entropy loss function: L1=-ylog(p)-(1-y)log(1-p) Among them, y is the true label of technical feasibility, y=0 represents infeasible, y=1 represents feasible, and p is the technical feasibility probability predicted by the model; The resource requirement loss is expressed using the mean square error loss function: Among them, y i is the actual resource demand, is the resource requirement predicted by the model, n is the number of samples; The risk assessment loss is expressed using the cross entropy loss function: Where C is the number of risk levels, b i is the real risk level label, p i is the probability of risk level i predicted by the model.

7. A full-process order lifecycle management method according to claim 6, characterized in that: The training of the multi-task neural network model adopts a dynamic weighting method, and dynamically adjusts the weight according to the learning situation of each task at different training stages, which specifically includes the following steps: Initialize weights: At the beginning of training, assign initial weights to the loss function of each task Weight adjustment during training: After each training cycle, calculate the loss change rate of each task in, represents the loss of the i-th task in the t-th training cycle, and the loss change rate of the first training cycle is initialized to a preset value; Adjust the weight of each task based on the loss change rate: in, is the weight of the i-th task at the t-th training cycle, α is a hyperparameter used to control the magnitude of the weight adjustment, and m is the number of tasks.

8. A full-process order lifecycle management system, characterized in that: For implementing the method according to any one of claims 1 to 7, the system comprises: Information acquisition and work order creation module: obtain user information and user needs, and create work orders; Pre-sales plan generation module: allocates work orders according to preset rules, supplements relevant information, generates pre-sales plans, and updates the work order status to pending evaluation; Evaluation module: automatically evaluates the technical feasibility, resource requirements and risks of the pre-sales solution, and generates an evaluation report. If the evaluation conclusion in the evaluation report is passed, the R&D progress synchronization module is called and the work order status is updated to under development. If the evaluation conclusion is rejected, the rejection reason is generated, and the pre-sales solution generation module is called to modify the pre-sales solution, and the work order status is updated to pre-sales modification; R&D progress synchronization module: Conduct project development based on pre-sales solutions and evaluation reports, update the R&D progress in real time during the R&D process, upload the R&D results after the R&D is completed, and update the work order status to pending testing; Testing module: Tests R&D results based on automated test scripts and automatically generates test reports. If the test passes, uploads the test report and updates the work order status to pending delivery. If the test fails, calls the R&D progress synchronization module to restart R&D and update the progress. Delivery module: Generates a delivery list based on R&D results and test reports, automatically sends it to the user, and updates the work order status to delivered; Feedback adjustment module: obtains user feedback issues and automatically classifies them according to the issues. If the classification result is a problem type, the R&D progress synchronization module is called to re-develop the issues based on the feedback. If the classification result is a demand type, the pre-sales solution generation module is called to regenerate the pre-sales solution based on the feedback. If all issues have been resolved or no issues exist, the work order status is updated to completed.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images