Design method and device for double-extraction steam turbine of ultra-supercritical unit

Abstract

The invention provides a design method and device for a double-extraction steam turbine of an ultra-supercritical unit, belongs to the field of double-extraction steam turbines, and is used for solving the problems that the design of the double-extraction steam turbine of the ultra-supercritical unit in the related technology is complex in work, difficult to comprehensively consider, low in efficiency and poor in effect. According to the method, the recommended design scheme can be intelligently analyzed and determined through simple design task description expressed by a natural language, the design process is greatly simplified, the design efficiency is improved, in the design process, the consideration directions of all parameter design dimensions, dimension performance scores of multiple performance dimensions, scheme cost values and the like are considered, and the design efficiency is improved. And the selectable design schemes are comprehensively determined as the recommended design schemes in combination with historical design records, the design consideration is comprehensive, the design process is intelligent and reasonable, and the finally determined recommended design scheme has extremely high reference value and is beneficial to improving the design effect.

Description

Technical Field

[0001] This application relates to the field of double-extraction steam turbines, and particularly to a design method and device for double-extraction steam turbines in ultra-supercritical units. Background Art

[0002] In today's energy and power field, thermal power generation, as one of the main power generation forms, continues to play a key role. Ultra-supercritical units have become an important direction for the development of thermal power generation technology due to their advantages of significantly improving the thermal efficiency of the unit, reducing coal consumption, and reducing pollutant emissions, and have been widely applied and promoted globally.

[0003] The double-extraction steam turbine in an ultra-supercritical unit, as a core component in the unit system, plays a crucial role in the safe and stable operation of the entire unit and the energy utilization efficiency. The double-extraction steam turbine realizes external heat supply or industrial steam supply by reasonably extracting steam at different pressure levels in the steam turbine, while ensuring the efficient progress of the power generation process. This design concept aims to improve the comprehensive utilization efficiency of energy, meet various energy consumption requirements in the industrial production process, and the requirements of urban central heating, and conforms to the current development trend of energy cascade utilization.

[0004] With the increasingly diverse and complex demands in the power market, as well as the continuous improvement of the requirements for energy utilization efficiency and unit operation stability, there are more and more dimensions to consider in the design of double-extraction steam turbines in ultra-supercritical units. Based on this, the difficulties in designing double-extraction steam turbines in ultra-supercritical units not only come from the fact that it is often difficult to accurately predict the application situation of the design scheme based on empirical formulas and simplified thermal calculation models at the design level, but also from the fact that designers need to spend time and effort to determine the requirements for all dimensions to be considered for double-extraction steam turbines in ultra-supercritical units based on a certain degree of professionalism, which brings great difficulties and challenges to the design work. Summary of the Invention

[0005] This application provides a design method and device for double-extraction steam turbines in ultra-supercritical units, which is conducive to improving the design efficiency, simplifying the design work, and improving the design effect in the design work of double-extraction steam turbines in ultra-supercritical units, in order to cope with the difficulties and challenges in the design work.

[0006] In the first aspect, this application provides a design method for double-extraction steam turbines in ultra-supercritical units. The method includes: Obtain a design task description of the double-extraction steam turbine, where the design task description is expressed in natural language; Determine an optional design scheme according to the design task description, where the optional design scheme includes multiple design parameter dimensions and the dimension parameter values of each design parameter dimension; Input the optional design solution into a pre-built cost prediction model and performance evaluation models for multiple performance dimensions to obtain the corresponding solution cost value and the dimensional performance scores for each performance dimension; Determine the solution recommendation degree of the optional design solution by combining the solution cost value, the dimensional performance scores of all performance dimensions, and the pre-obtained historical design records; Determine the recommended design solution from the optional design solutions according to the solution recommendation degree.

[0007] By adopting the above technical solution, the designer only needs to input the design task description expressed in natural language, and this algorithm model can automatically analyze the optional design solutions containing multiple design parameter dimensions, automatically comprehensively evaluate the performance of the optional design solutions in all aspects, and finally determine the solution recommendation degree of the optional design solution by combining the historical design records, and finally determine the appropriate recommended design solution, which is beneficial to reducing the workload of the designer, simplifying the design work, and fully considering various design dimensions and historical design records is beneficial to greatly improving the design effect.

[0008] Further, the historical design records carry time stamps, and the historical design records include design description records and selected design solutions; The determining the solution recommendation degree of the optional design solution by combining the solution cost value, the dimensional performance scores of all performance dimensions, and the pre-obtained historical design records includes: Determine the recommendation degree base value based on the solution cost value and the dimensional performance scores of all performance dimensions; Screen similar design records from the historical design records according to the design task description, and the similarity between the design description records of the similar design records and the design task description is higher than the description similarity threshold; Analyze the recommendation degree influence coefficient according to the similar design records of the optional design solution; Calculate the solution recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient.

[0009] Further, the determining the recommendation degree base value based on the solution cost value and the dimensional performance scores of all performance dimensions includes: Let the solution cost value be c, the dimensional performance score of the i-th dimension be , and the recommendation degree base value be , then ; In the formula, is the attention coefficient constant preset for the i-th performance dimension.

[0010] Further, the analyzing the recommendation degree influence coefficient according to the similar design records of the optional design solution includes: Let the selected design solution of the \(i\)-th similar design record in chronological order of timestamps be , the timestamp be , the alternative design solutions be , the current moment be , and the recommendation influence coefficient be \(k\). Then ; ; ; ; ; ; wherein, is the slope extraction function, which represents the slope of the fitting line obtained by linearly regressing the coordinate points of the similar design records with the abscissa being and the ordinate being , and is the pre-obtained basic influence coefficient.

[0011] Furthermore, calculating the scheme recommendation degree according to the recommendation base value and the recommendation influence coefficient includes: Let the recommendation base value be , the recommendation influence coefficient be \(k\), and the scheme recommendation degree be \(z\). Then ; wherein, \(A\) is the pre-obtained scheme confidence level, the dimension parameter value of the design parameter dimension carries the dimension confidence level based on the design task description, and the scheme confidence level is positively correlated with all dimension confidence levels.

[0012] In a second aspect, the present application provides a double-extraction steam turbine design device for an ultra-supercritical unit. The device includes: an acquisition module, configured to acquire a design task description of the double-extraction steam turbine, where the design task description is expressed in natural language; an understanding module, configured to determine alternative design solutions according to the design task description, where the alternative design solutions include multiple design parameter dimensions and dimension parameter values of each design parameter dimension; an analysis module, configured to input the alternative design solutions into a pre-constructed cost prediction model and performance evaluation models of multiple performance dimensions to obtain corresponding scheme cost values and dimension performance scores of each performance dimension; a calculation module, configured to determine the scheme recommendation degree of the alternative design solutions by combining the scheme cost value, the dimension performance scores of all performance dimensions, and pre-obtained historical design records; and a recommendation module, configured to determine a recommended design solution from the alternative design solutions according to the scheme recommendation degree.

[0013] Further, the historical design record carries a time stamp, and the historical design record includes a design description record and a selected design scheme; The calculation module is further configured that determining a scheme recommendation degree of an optional design scheme based on the combined scheme cost value, the dimension performance scores of all performance dimensions, and the pre-acquired historical design record includes: Determining a recommendation degree base value based on the scheme cost value and the dimension performance scores of all performance dimensions; Filtering similar design records from the historical design records according to the design task description, where the similarity between the design description record of the similar design record and the design task description is higher than a description similarity threshold; Analyzing a recommendation degree influence coefficient according to the similar design records of the optional design scheme; Calculating the scheme recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient.

[0014] Further, the calculation module is further configured that determining the recommendation degree base value based on the scheme cost value and the dimension performance scores of all performance dimensions includes: Let the scheme cost value be c, the dimension performance score of the i-th dimension be , and the recommendation degree base value be , then ; In the formula, is a concern coefficient constant preset for the i-th performance dimension.

[0015] Further, the calculation module is further configured that analyzing the recommendation degree influence coefficient according to the similar design records of the optional design scheme includes: Let the selected design scheme of the i-th similar design record in chronological order of time stamps be , the time stamp be , the optional design scheme be , the current moment be , and the recommendation influence coefficient be k, then ; ; ; ; ; ; In the formula, is a slope extraction function, which represents that the abscissa is , and the ordinate is The slope of the fitting line obtained by linear regression of the coordinate points of the similar design record, is the basic influence coefficient to be pre-obtained.

[0016] Furthermore, the calculation module is further configured that calculating the solution recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient includes: Let the recommendation degree base value be , the recommendation influence coefficient be k, and the solution recommendation degree be z, then ; In the formula, A is the solution confidence degree to be pre-obtained, the dimension parameter value of the design parameter dimension carries the dimension confidence degree based on the design task description, and the solution confidence degree is positively correlated with all dimension confidence degrees.

[0017] In summary, the present application at least includes the following beneficial effects: A design method and device for a double-extraction steam turbine of an ultra-supercritical unit are provided, which can simplify the design work, improve the design efficiency, and optimize the design effect.

[0018] It should be understood that the content described in the invention content part is not intended to limit the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Combined with the drawings and referring to the following detailed description, the above and other features, advantages, and aspects of the embodiments of the present application will become more obvious. In the drawings, the same or similar reference numerals represent the same or similar elements, where: Figure 1 shows a flowchart of a design method for a double-extraction steam turbine of an ultra-supercritical unit in an embodiment of the present application; Figure 2 shows a block diagram of a design device for a double-extraction steam turbine of an ultra-supercritical unit in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0021] In addition, the term "and / or" in this text merely describes the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this text generally indicates that the associated objects before and after are in an "or" relationship.

[0022] The present application provides a design method and device for a double-extraction steam turbine of an ultra-supercritical unit, which can simplify the design work, improve the design efficiency, and optimize the design effect based on an intelligent algorithm model.

[0023] In a first aspect, an embodiment of the present application discloses a design method for a double-extraction steam turbine of an ultra-supercritical unit. This method can be specifically executed by a server Figure 1 The flowchart of a design method for a double-extraction steam turbine of an ultra-supercritical unit in an embodiment of the present application is shown.

[0024] Referring to Figure 1 , the method specifically includes the following steps: S110: Obtain a design task description of the double-extraction steam turbine, and the design task description is expressed in natural language.

[0025] Regarding the method for obtaining the design task description, it can be obtained based on text input, audio input, video input, or human-computer dialogue, etc. The original file of the design task description can be a design contract or a design project text, the audio and video of the communication record with the customer, or the dialogue process record between the designer and a general or proprietary large model. After performing key information extraction, collation, and manual review on the text information contained in the original file based on known information processing technologies, a design task description expressed in natural language is formed. Of course, the design task description can also be the direct natural language input of the designer.

[0026] S120: Determine an optional design scheme according to the design task description, and the optional design scheme includes multiple design parameter dimensions and the dimension parameter values of each design parameter dimension.

[0027] The method of this step is implemented based on natural language processing technology and a decision tree pre-constructed based on industry consensus. Among them, natural language processing technology is used to parse the design task description, and the decision tree is used to determine the optional design scheme based on the parsing result.

[0028] Use natural language processing technology to deeply analyze the natural language description of the design task and accurately extract key information, including but not limited to: Performance indicators: For example, when designing a double-extraction steam turbine for an ultra-supercritical unit with an expected power generation of 600 MW, a stable high-pressure extraction steam pressure of 4 - 5 MPa, and a low-pressure extraction steam flow rate to meet the heating demand of 150 t / h, accurately extract core performance points such as "600 MW power generation", "4 - 5 MPa high-pressure extraction steam pressure", and "150 t / h low-pressure extraction steam flow rate".

[0029] Operating conditions: If it is mentioned that "this steam turbine will be installed in a high-altitude and cold area with an environmental temperature as low as -20°C and a relatively high dust content in the air", then carefully extract key details of operating conditions such as "high altitude", "low temperature", and "dust environment". These factors will significantly affect the material selection of the equipment, the design of the cooling system, and the formulation of protection measures.

[0030] Reliability and maintenance requirements: For example, "it is required to have extremely high reliability, with an average trouble-free operation time of not less than 5000 hours, and the maintenance should be convenient and efficient, with a single maintenance duration of not more than 48 hours". Clearly grasp key requirements such as "high reliability" and "short maintenance duration", which will play a key role in component selection, structural design optimization, and maintenance strategy planning.

[0031] Regarding the construction of the decision tree, the main content is the setting of decision tree nodes and branches. For example, taking the power generation as the root node: According to industry experience, for different power generation ranges (such as 300 - 500 MW, 500 - 800 MW, etc.), determine the approximate value range or alternative solution branches of subsequent relevant parameters. For example, when the power generation is around 600 MW, the main steam pressure may be more appropriate within the range of 28 - 32 MPa, which forms a branch in the dimension of the main steam pressure.

[0032] Consider the extraction parameter branches: For extraction parameters such as high-pressure extraction steam pressure and low-pressure extraction steam flow rate, further subdivide the branches. For example, when the high-pressure extraction steam pressure is 4 - 5 MPa, it may correspond to a specific number of stages range (such as 8 - 10 stages) of the intermediate-pressure cylinder and a range of extraction port positions (such as after the 4th - 6th stages of the high-pressure cylinder). These branches are constructed based on the parameter correlation relationships determined by a large number of actual cases and theoretical calculations in the industry.

[0033] Combine the operating environment branches: When the operating environment is high-altitude, cold, and dusty, in the material selection branch, it may point to special alloy materials that are resistant to high temperature, low temperature, and dust erosion; in the cooling system branch, it may tend to adopt a cooling method that combines air cooling and water cooling and has a design to prevent dust blockage. Each branch has corresponding parameter value ranges or specific selection suggestions, which are all based on the summary and refinement of the operating experience of equipment in similar environments in the industry.

[0034] Of course, considering the reliability of the decision tree, confidence levels can also be added to the decision tree branches. The confidence levels are achieved based on data statistics and empirical evaluations. Specifically, for example, through statistical analysis of historical data from a large number of similar design projects in the industry and combined with expert experience judgments, confidence levels are assigned to each branch of the decision tree. For example, for a plan where the power generation is 600MW and the high-pressure extraction steam pressure is 4 - 5MPa, and the medium-pressure cylinder selects a 9-stage plan. If 40 out of the past 50 similar projects have adopted a similar configuration and operated well, then the confidence level of this branch can be set to 0.8 (40 / 50). For some emerging technologies or parameter combinations with less application, the confidence levels may be relatively low. For example, for a plan using a certain new high-temperature resistant coating, due to fewer actual application cases, its confidence level may only be 0.3 - 0.5.

[0035] Based on the key information processed by natural language processing technology and the constructed decision tree, it is convenient for the decision tree to determine optional solutions. For example, starting from the root node of the decision tree, according to the key information extracted from the design task, traverse along the corresponding branches, combine the values of each parameter dimension, and generate a complete optional design solution. For example, a solution may be: the main steam pressure is 30MPa (confidence level 0.7), the main steam temperature is 620°C (confidence level 0.8), the high-pressure cylinder has 10 stages (confidence level 0.85), the low-pressure cylinder has 6 stages (confidence level 0.8), the high-pressure extraction steam port is after the 5th stage (confidence level 0.9), the extraction steam pipeline diameter is 0.4m (confidence level 0.75), etc. Each parameter dimension carries its corresponding confidence level label, and these confidence level labels reflect the reliability of the parameter value under the current design task and industry experience, providing an important reference basis for subsequent solution evaluation and optimization.

[0036] Based on the above content, the optional design solutions can be determined. The optional design solutions are expressed as multi-dimensional vectors and can carry the confidence levels of each parameter design dimension.

[0037] S130: Input the said optional design solution into a pre-constructed cost prediction model and performance evaluation models for multiple performance dimensions to obtain the corresponding solution cost value and the dimension performance scores for each performance dimension.

[0038] Since the optional design solution is essentially the dimension parameter values of all parameter design dimensions, reflecting the overall design requirements for each dimension of the double-extraction steam turbine. Based on these requirements, as well as industry experience data, design experience records, etc., the cost of the double-extraction steam turbine and its performance performance in multiple aspects can be predicted. The pre-trained and constructed cost prediction model is used to analyze the solution cost value of the optional design solution based on the optional design solution, and the pre-constructed performance evaluation model is used to analyze the dimension performance score of a certain performance dimension based on the optional design solution.

[0039] Regarding the cost prediction model, in one example, it can be manifested as substituting the alternative design solutions into the cost prediction model , and calculating multiple cost components, such as material cost , manufacturing process cost , equipment selection cost , operation and maintenance cost , etc. The comprehensive cost score is calculated by comparing each cost component with a pre-set cost target value or reference value, and according to the cost score conversion function . For example, assuming the total target cost is , and the actual total cost is , then: ; For example, the conversion function is: ; Based on the above content, the cost value of the alternative design solution can be analyzed and obtained.

[0040] Regarding the performance evaluation model, specifically, for example, the power generation efficiency evaluation model, the air extraction performance evaluation model, and the reliability evaluation model.

[0041] Regarding the performance score in terms of power generation efficiency dimension , substitute the alternative design solution into the power generation efficiency evaluation model , calculate the power generation efficiency prediction value , and then according to the pre-set power generation efficiency target value and the performance score conversion function , calculate the performance score in terms of power generation efficiency dimension: ; For example, assuming the conversion function is a linear function: ; Regarding the air extraction performance evaluation model ( ), substitute the alternative design solution into the steam extraction performance evaluation model , calculate indicators such as the high-pressure steam extraction pressure deviation , the low-pressure steam extraction flow deviation , etc., and then determine the performance score in terms of steam extraction performance dimension according to the deviation and the performance score conversion function . For example, for the high-pressure steam extraction pressure deviation, the conversion function is: ; In the formula, is the preset high-pressure extraction pressure. Similarly, the low-pressure extraction performance score can be obtained and will not be elaborated here. Assume that the performance score of the extraction steam performance dimension is obtained by weighted average of the high-pressure extraction and low-pressure extraction performance scores, and the weights are , then: ; Regarding the reliability evaluation model , substitute the alternative design scheme into the reliability evaluation model , calculate the reliability index , and then determine the reliability dimension performance score according to the conversion function between the reliability index and the performance score : ; For example, assume that the value range of the reliability index is between 0 and 1, and the conversion function is: ; And so on, calculate the performance scores of other dimensions and will not list them one by one.

[0042] S140: Determine the scheme recommendation degree of the alternative design scheme by combining the scheme cost value, the dimension performance scores of all performance dimensions, and the pre-obtained historical design records.

[0043] The method of this step specifically includes: determining the recommendation degree base value based on the scheme cost value and the dimension performance scores of all performance dimensions; screening similar design records in the historical design records according to the design task description, and the description similarity between the design description record of the similar design record and the design task description is higher than the description similarity threshold; analyzing the recommendation degree influence coefficient according to the similar design records of the alternative design scheme; calculating the scheme recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient.

[0044] In an example, determining the recommendation degree base value based on the scheme cost value and the dimension performance scores of all performance dimensions includes: setting the scheme cost value as c, the performance score of the i-th dimension as , and the recommendation degree base value as , then , where is the attention coefficient constant preset for the i-th performance dimension.

[0045] In an example, analyzing the recommendation degree influence coefficient according to the similar design records of the alternative design scheme includes: setting the selected design scheme of the i-th similar design record in chronological order as , the time stamp as , the alternative design scheme as , and the current moment as , if the recommended influence coefficient is k, then ; ; ; ; ; ; In the formula, is the slope extraction function, which represents the fitting line slope obtained by linear regression of the similarity design record coordinate points with the abscissa being and the ordinate being , and is the pre-obtained basic influence coefficient.

[0046] In an example, calculating the scheme recommendation degree according to the recommendation degree base value and the recommendation influence coefficient includes: setting the recommendation degree base value as , the recommendation influence coefficient as k, and the scheme recommendation degree as z, then , where A is the pre-obtained scheme confidence level, the dimension parameter value of the design parameter dimension carries the dimension confidence level based on the design task description, and the scheme confidence level is positively correlated with all dimension confidence levels. Specifically, for example, it is equal to the sum or average of all dimension confidence levels.

[0047] It should be understood that on the basis of being negatively correlated with the scheme cost value and positively correlated with each dimension performance score, the foregoing recommendation degree base value can also be considered to be designed as other consensus structures. The specifically designed formula structure in this step of the method can ensure that the result range of the recommendation degree base value is between 0 and 1. In addition, regarding the recommendation influence coefficient, it can also be only associated with one or two of them to adaptively adjust the formula structure. Correspondingly, when the formula structures of the recommendation degree base value and the recommendation influence coefficient change, the formula structure of the scheme recommendation degree can be adaptively adjusted. In short, the specific determination process of the scheme recommendation degree can also be implemented in other forms, which will not be listed one by one here.

[0048] S150: Determine the recommended design scheme among the optional design schemes according to the scheme recommendation degree.

[0049] In an example, after the scheme recommendation degrees of all optional design schemes are determined, one or a specified number of optional design schemes with the highest recommendation degree can be considered as the recommended design schemes and recommended to the user.

[0050] In another example, when multiple recommended design solutions are required, in order to enrich the content of the recommended design solutions, the optional design solutions can also be sorted in descending order according to the scheme recommendation degree to form a set of optional solutions. First, select the first optional design solution as the recommended design solution, and make the set of optional solutions discard the selected optional design solution and the optional design solutions whose vector similarity with the selected optional design solution is higher than the preset recommended similarity threshold. Repeat the above actions until the number of selected optional design solutions reaches the specified number. Then, use the selected optional design solutions as the recommended design solutions.

[0051] The user can select the selected design solution of the double-extraction steam turbine from the recommended design solutions, or adjust a recommended design solution to form a new design solution as the selected design solution. Of course, the user can also discard all the recommended design solutions and construct the selected design solution by himself / herself.

[0052] Based on the above, this method can intelligently analyze and determine the recommended design solutions through the design task description expressed in simple natural language, greatly simplifying the design process, facilitating the improvement of design efficiency. Moreover, in the design process, not only all parameter design dimensions, dimensional performance scores of multiple performance dimensions, and scheme cost values and other consideration directions are considered, but also historical design records are combined to comprehensively determine the optional design solutions as the recommended design solutions. The design is comprehensively considered, the design process is intelligent and reasonable, and the finally determined recommended design solutions have extremely high reference value, which is beneficial to improving the design effect.

[0053] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to the embodiments of this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0054] In a second aspect, an embodiment of this application discloses a design device for a double-extraction steam turbine of an ultra-supercritical unit. This device can be implemented as a server for design work or be included in a server.

[0055] Figure 2 The block diagram of a design device for a double-extraction steam turbine of an ultra-supercritical unit in an embodiment of this application is shown.

[0056] Refer to Figure 2 , the device specifically includes: An acquisition module 210, configured to acquire a design task description of a double-extraction steam turbine, where the design task description is expressed in natural language; An understanding module 220, configured to determine alternative design solutions according to the design task description, where the alternative design solutions include multiple design parameter dimensions and dimension parameter values for each design parameter dimension; An analysis module 230, configured to input the alternative design solutions into a pre-constructed cost prediction model and performance evaluation models for multiple performance dimensions to obtain corresponding solution cost values and dimension performance scores for each performance dimension; A calculation module 240, configured to determine a solution recommendation degree of the alternative design solutions by combining the solution cost value, the dimension performance scores of all performance dimensions, and pre-acquired historical design records; and A recommendation module 250, configured to determine a recommended design solution from the alternative design solutions according to the solution recommendation degree.

[0057] Further, the historical design records carry timestamps, and the historical design records include design description records and selected design solutions; The calculation module 240 is further configured such that the determining the solution recommendation degree of the alternative design solutions by combining the solution cost value, the dimension performance scores of all performance dimensions, and the pre-acquired historical design records includes: Determining a recommendation degree base value based on the solution cost value and the dimension performance scores of all performance dimensions; Filtering similar design records from the historical design records according to the design task description, where the similarity between the design description records of the similar design records and the design task description is higher than a description similarity threshold; Analyzing a recommendation degree influence coefficient according to the similar design records of the alternative design solutions; Calculating the solution recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient.

[0058] Further, the calculation module 240 is further configured such that the determining the recommendation degree base value based on the solution cost value and the dimension performance scores of all performance dimensions includes: Let the solution cost value be c, the dimension performance score of the i-th dimension be , and the recommendation degree base value be , then ; In the formula, is a predefined attention coefficient constant for the i-th performance dimension.

[0059] Further, the calculation module 240 is further configured such that the analyzing the recommendation degree influence coefficient according to the similar design records of the alternative design solutions includes: Let the selected design solution of the i-th similar design record in chronological order of timestamps be , and the timestamp be , the alternative design is , the current time is , if the recommended influence coefficient is k, then ; ; ; ; ; ; In the formula, is the slope extraction function, which represents the fitting line slope obtained by linear regression of the coordinate points of the similar design record with the abscissa being and the ordinate being , and is the pre-obtained basic influence coefficient.

[0060] Furthermore, the calculation module 240 is further configured such that the calculating the scheme recommendation degree according to the recommendation base value and the recommendation influence coefficient includes: Let the recommendation base value be , the recommendation influence coefficient be k, and the scheme recommendation degree be z, then ; In the formula, A is the pre-obtained scheme confidence level, the dimension parameter value of the design parameter dimension carries the dimension confidence level based on the design task description, and the scheme confidence level is positively correlated with all dimension confidence levels.

[0061] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the described device can refer to the corresponding process in the foregoing method embodiment, and will not be elaborated herein.

[0062] In summary, the present application at least includes the following beneficial effects: A design method and device for a double-extraction steam turbine of an ultra-supercritical unit are provided, which can simplify the design work, improve the design efficiency, and optimize the design effect.

[0063] The above description is only the preferred embodiment of the present application and the description of the applied technical principle. Those skilled in the art should understand that the disclosed scope in the present application is not limited to the technical solution formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the foregoing disclosed concept. For example, the technical solution formed by mutually replacing the above features with the (but not limited to) technical features having similar functions disclosed in the present application.

Claims

1. A design method for a double-extraction steam turbine of an ultra-supercritical unit, characterized in that: include: Obtain a design task description of a double-extraction steam turbine, wherein the design task description is expressed in a natural language; Determine an optional design solution according to the design task description, wherein the optional design solution includes multiple design parameter dimensions and a dimension parameter value of each design parameter dimension; Input the optional design scheme into a pre-built cost prediction model and a performance evaluation model of multiple performance dimensions to obtain a corresponding scheme cost value and a dimensional performance score of each performance dimension; Determine the recommendation degree of the optional design solution by combining the solution cost value, the dimension performance scores of all performance dimensions, and the pre-acquired historical design records; A recommended design scheme is determined from the optional design schemes according to the recommendation degree of the scheme.

2. The method according to claim 1, characterized in that The historical design record carries a timestamp, and the historical design record includes a design description record and a selected design solution; The scheme recommendation degree of the optional design scheme is determined by combining the scheme cost value, the dimensional performance scores of all performance dimensions, and the pre-acquired historical design records, including: Determine a base value of recommendation based on the cost value of the solution and the dimensional performance scores of all performance dimensions; Screening similar design records from historical design records according to the design task description, wherein the description similarity between the design description record of the similar design record and the design task description is higher than a description similarity threshold; Analyze the recommendation influence coefficient according to the similar design records of the optional design solutions; The recommendation degree of the scheme is calculated according to the recommendation degree base value and the recommendation degree influence coefficient.

3. The method according to claim 2, characterized in that Determining the recommendation degree base value based on the solution cost value and the dimension performance scores of all performance dimensions includes: Assume that the solution cost value is c and the performance score of the i-th dimension is The recommended base value is ,but ； In the formula, is the preset attention coefficient constant relative to the i-th performance dimension.

4. The method according to claim 2, characterized in that: The recommendation degree influence coefficient of the similar design record analysis according to the optional design solution includes: Suppose the selected design scheme of the i-th similar design record in the order of timestamp is , timestamp is , the optional design options are , the current time is , the recommended influence coefficient is k, then ； ； ； ； ； ； In the formula, is the slope extraction function, which represents the horizontal coordinate , the vertical axis is The slope of the fitted line obtained by linear regression of the coordinate points of the similar design record is, is the basic influence coefficient obtained in advance.

5. The method according to claim 2, characterized in that: Calculating the recommendation degree of the scheme according to the recommendation degree base value and the recommendation degree influence coefficient comprises: Assume the recommended base value is , the recommended influence coefficient is k, and the recommended degree of the scheme is z, then ； In the formula, A is the pre-acquired solution confidence, the dimension parameter value of the design parameter dimension carries the dimension confidence based on the design task description, and the solution confidence is positively correlated with all dimension confidences.

6. A design device for a double extraction steam turbine of an ultra-supercritical unit, characterized in that: include: An acquisition module (210) is used to acquire a design task description of a double-extraction steam turbine, wherein the design task description is expressed in a natural language; An understanding module (220), configured to determine an optional design solution according to the design task description, wherein the optional design solution includes a plurality of design parameter dimensions and a dimension parameter value of each design parameter dimension; An analysis module (230) is used to input the optional design scheme into a pre-built cost prediction model and a performance evaluation model of multiple performance dimensions to obtain a corresponding scheme cost value and a dimensional performance score of each performance dimension; A calculation module (240) is used to determine the recommendation degree of the optional design solution by combining the solution cost value, the dimension performance scores of all performance dimensions, and the pre-acquired historical design records; as well as The recommendation module (250) is used to determine a recommended design solution from the optional design solutions according to the solution recommendation degree.

7. The device according to claim 6, characterized in that The historical design record carries a timestamp, and the historical design record includes a design description record and a selected design solution; The calculation module (240) is further configured to determine the recommendation level of the optional design solution by combining the solution cost value, the dimension performance scores of all performance dimensions, and the pre-acquired historical design records, including: Determine a base value of recommendation based on the cost value of the solution and the dimensional performance scores of all performance dimensions; Screening similar design records from historical design records according to the design task description, wherein the description similarity between the design description record of the similar design record and the design task description is higher than a description similarity threshold; Analyze the recommendation influence coefficient according to the similar design records of the optional design solutions; The recommendation degree of the scheme is calculated according to the recommendation degree base value and the recommendation degree influence coefficient.

8. The device according to claim 7, characterized in that The calculation module (240) is further configured to determine the recommendation degree base value based on the solution cost value and the dimension performance scores of all performance dimensions, including: Assume that the solution cost value is c and the performance score of the i-th dimension is The recommended base value is ,but ； In the formula, is the preset attention coefficient constant relative to the i-th performance dimension.

9. The device according to claim 8, characterized in that The calculation module (240) is further configured such that the analysis of the recommendation degree influence coefficient of the similar design records according to the optional design solutions includes: Suppose the selected design scheme of the i-th similar design record in the order of timestamp is , timestamp is , the optional design options are , the current time is , the recommended influence coefficient is k, then ； ； ； ； ； ； In the formula, is the slope extraction function, which represents the horizontal coordinate , the vertical axis is The slope of the fitted line obtained by linear regression of the coordinate points of the similar design record is, is the basic influence coefficient obtained in advance.

10. The device according to claim 7, characterized in that The calculation module (240) is further configured to calculate the scheme recommendation degree according to the recommendation degree base value and the recommendation degree influence coefficient, including: Assume the recommended base value is , the recommended influence coefficient is k, and the recommended degree of the scheme is z, then ； In the formula, A is the pre-acquired solution confidence, the dimension parameter value of the design parameter dimension carries the dimension confidence based on the design task description, and the solution confidence is positively correlated with all dimension confidences.

Images