CN-122022871-A - Tire changer sales strategy self-adaptive generation system based on reinforcement learning

CN122022871ACN 122022871 ACN122022871 ACN 122022871ACN-122022871-A

Abstract

The invention relates to the technical field of tire changer sales, in particular to a tire changer sales strategy self-adaptive generation system based on reinforcement learning, which comprises a supply and demand characteristic sensing module for dividing areas and analyzing price trend generation characteristics, a conversion fluctuation identification module for identifying conversion efficiency deviation positions, the state operation association module analyzes the multidimensional attribute to establish mapping, the interaction flow arrangement module arranges operations according to the state sequence, and the interaction behavior self-adaptive adjustment module rearranges the operations according to feedback positioning deviation and selects alternative operations to obtain a self-adaptive interaction strategy generation result. According to the invention, the market state is structured through multidimensional disassembly and consistency characterization of the formation in the sales process, abnormal limitation is converted to the position, and the positioning is directional through associated business constraint, stable mapping is established between the flow state and the operation, so that interactive predictability basis is given, and the sales behavior is kept consistent and adjustable through feedback rearrangement of the operation sequence, so that supply and demand identification definition, interactive stability and strategy adaptation performance are enhanced.

Inventors

ZHANG YANJIAO
HE ZHONGXIN
Jin Pingchuan

Assignees

吉林农业科技学院

Dates

Publication Date: 20260512
Application Date: 20260304

Claims (8)

1. A reinforcement learning-based tire changer sales strategy adaptive generation system, the system comprising: the supply and demand feature perception module is used for acquiring service node information of a sales area, dividing the sales area, distributing area node numbers, collecting transaction data, generating trend vectors, analyzing trend consistency, distinguishing supply and demand performance types and generating a supply and demand data feature area result; The conversion fluctuation identification module is used for obtaining the interactive operation conversion efficiency of each regional node according to the regional node number given to the supply and demand characteristic label in the supply and demand data characteristic regional result, extracting response time delay and protocol fluctuation, identifying the conversion efficiency deviation position, outputting the operation position number and generating a conversion fluctuation position result; the state operation association module extracts the client attribute, time delay, protocol change, feedback signal and lead time associated with the corresponding operation according to the operation position number in the conversion fluctuation position result, analyzes the relation with the interactive operation, establishes the corresponding description of the flow state and the interactive operation, and generates a state operation mapping result; and the interactive flow arranging module acquires the actual sequence of the flow states in the flow execution process according to the corresponding description of the flow states and the interactive operations in the state operation mapping result, sequentially selects and arranges the interactive operation records to form continuous propelling description, and generates an interactive operation sequence result.
2. The self-adaptive generation system of a tire changer sales strategy based on reinforcement learning according to claim 1, wherein the supply and demand data characteristic region result comprises a region volume change level, a region price change direction type, a region supply and demand tension degree division and a region volume fluctuation range characteristic, the conversion fluctuation position result comprises a conversion abnormal occurrence sequence number, a conversion efficiency reduction section and a conversion unstable operation point, the state operation mapping result comprises a flow state type set, a state corresponding interactive operation type and a state matching condition element, and the interactive operation sequence result comprises an interactive operation arrangement sequence, an operation connection structure form and a flow propulsion rhythm description.
3. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning of claim 1, wherein the supply and demand characteristic sensing module comprises a regional node marking sub-module, a trend vector construction sub-module and a supply and demand type marking sub-module; The regional node dividing sub-module is used for acquiring service execution node information corresponding to each sales region in the process of selling by the tire removing machine, dividing the sales region according to service coverage and sales execution boundary, setting regional node numbers for each sales region, collecting the information of the number of deals and the change direction of the price of the deals in the region corresponding to the regional node numbers, and generating a node number mapping list; The trend vector construction submodule calls the collected number of the deals and the change direction information of the deals price in the node number mapping list, arranges the orders of the numbers of the deals in different time periods under the same area node number, sorts the arrangement order of the change direction information of the deals price, forms continuous records respectively, generates the change vector of the number of the deals and the change vector of the deals price according to time sequence, and generates a trend vector sorting list after summarizing and sorting; And the supply and demand type labeling sub-module is used for identifying trend of the change of the number of the achievement and the price in the time advancing process based on two types of vector sequence information corresponding to the node numbers of each region in the trend vector sequencing list, carrying out label identification processing on the node numbers of the regions according to the trend representation condition, and correspondingly matching the node numbers after the identification processing with the label information to generate a supply and demand data characteristic region result.
4. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning of claim 1, wherein the transformation fluctuation identification module comprises a node tag extraction sub-module, an interaction information collection sub-module and an offset position identification sub-module; the node label extraction submodule acquires all regional node numbers given with supply and demand characteristic labels in the supply and demand data characteristic region result, screens out node number contents with clear label identifications from a regional node number list, performs field matching on the regional node numbers with the label identifications and the interactive operation record numbers corresponding to the service execution stage, sorts corresponding combined contents between the regional node numbers and the interactive operation record numbers, and generates a node interactive record index table; The interactive information gathering sub-module is used for calling the interactive operation record numbers in the node interactive record index table, extracting conversion efficiency values, contact response delay information and settlement protocol change conditions corresponding to the current records from each interactive operation record, carrying out arrangement gathering on the conversion efficiency values, the contact response delay information and the settlement protocol change conditions according to the interactive operation record numbers, and establishing information groups according to the regional node numbers to generate a node interactive feature set; And the offset position identification sub-module judges whether the conversion efficiency change directions among the continuous records are consistent according to the conversion efficiency numerical value arrangement sequence corresponding to each group of interactive operation records in the node interactive feature set, screens out the interactive operation record numbers with inconsistent change directions, simultaneously performs joint screening treatment on the contact response time delay numerical value and the settlement protocol change numerical value on the screened interactive operation record numbers, screens out position numbers with conversion efficiency offset in the interactive sequence, and generates conversion fluctuation position results.
5. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning of claim 1, wherein the state operation association module comprises an operation record extraction sub-module, an attribute feature collection sub-module and a state type identification sub-module; An operation record extraction sub-module, which is used for obtaining operation position numbers listed in the conversion fluctuation position result, retrieving corresponding interactive operation record numbers from each operation position number, performing field correspondence through the operation record numbers and interactive execution records in the service data source, uniformly archiving the successfully matched interactive execution records, and generating an operation record number set; the attribute feature collection sub-module calls each interactive execution record in the operation record number set, extracts a client attribute feature value, a contact response delay value, a settlement protocol change state, a quotation feedback signal type and a lead time period corresponding to the interactive operation record, finishes structural arrangement of five items of extracted information by taking the operation record number as an index, classifies and combines the five items of extracted information according to fields, and generates an interactive attribute feature group; And the state type identification sub-module judges the distribution condition and the combination difference of attribute combination in the interactive operation sequence according to the client attribute characteristic value, the contact response delay value, the settlement protocol change state, the quotation feedback signal category and the delivery cycle time period corresponding to each operation record number in the interactive attribute characteristic group, classifies the operation record numbers with similar attribute combination content into the same flow state, sequentially establishes a matching structure between the interactive operation and the flow state according to the flow state sequence, and generates a state operation mapping result.
6. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning according to claim 1, wherein the interactive flow arrangement module comprises a state sequence acquisition sub-module, an operation record selection sub-module and a sequence arrangement generation sub-module; the state sequence obtaining sub-module is used for obtaining the corresponding description of the flow states and the interactive operation in the state operation mapping result, collecting the actual appearance sequence information of each flow state in the process of executing the flow, sorting and archiving the appearance sequence content of the flow states according to the time pushing sequence, recording the relation between the flow state numbers and the appearance sequence indexes, and generating a flow state sequence index table; The operation record selection sub-module is used for calling each flow state number in the flow state sequence index table, screening the content of the interactive operation record associated with the flow state number according to the interactive operation record number corresponding to the flow state in the state operation mapping result, carrying out number aggregation on the screened interactive operation record according to the flow state sequence index, and generating an interactive record index list; And the sequence arrangement generating sub-module judges whether the interactive operation records have missing positions in the index sequence according to the interactive operation record number arrangement sequence in the interactive record index list, sequentially links and arranges the interactive operation records with continuous numbers to form interactive operation pushing sequence description content, and generates an interactive operation sequence result.
7. The reinforcement learning-based tire changer sales strategy adaptive generation system of claim 1, further comprising: the interactive behavior self-adaptive adjustment module is used for acquiring conversion feedback data corresponding to the interactive operation advancing sequence in the actual execution process based on the interactive operation advancing sequence in the interactive operation sequence result, positioning the interactive operation deviating position, selecting a substitute interactive operation record, rearranging the interactive operation sequence and generating a self-adaptive interactive strategy generating result; The adaptive interaction strategy generation result comprises alternative interaction operation options, an operation sequence adjustment scheme and interaction strategy configuration parameters.
8. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning of claim 7, wherein the self-adaptive adjustment module of the interaction behavior comprises a feedback data extraction sub-module, an offset position positioning sub-module and a sequential rearrangement generation sub-module; The feedback data extraction sub-module acquires the interactive operation pushing sequence in the interactive operation sequence result, takes each interactive operation pushing number as an index, acquires conversion feedback data content corresponding to the actual execution process, and uniformly files the conversion feedback data and the interactive operation pushing numbers after corresponding to each conversion feedback data content to generate an operation feedback matching list; The offset position positioning sub-module is used for judging whether a position number with inconsistent feedback values and state descriptions exists in the advancing sequence according to the numerical content of conversion feedback data corresponding to each interactive operation advancing number in the operation feedback matching list and combining the interactive operation description corresponding to each flow state in the state operation mapping result, and executing screening and positioning on the position number to generate an advancing offset position number set; And the sequence rearrangement generating sub-module calls each number in the pushing offset position number set, acquires the interactive operation record corresponding to the number, screens the associated alternative interactive operation record content, replaces the abnormal position number in the original interactive operation pushing sequence with the corresponding alternative interactive operation record number, and then recombines the operation sequence to generate a self-adaptive interactive strategy generating result.

Description

Tire changer sales strategy self-adaptive generation system based on reinforcement learning Technical Field The invention relates to the technical field of tire changer sales, in particular to a tire changer sales strategy self-adaptive generation system based on reinforcement learning. Background The technical field of tire removing machine sales comprises the contents of mechanical equipment sales management, market policy optimization and the like, and the core of the technical field is that sales strategies are constructed, evaluated and adjusted by deep analysis and modeling of market demands, customer behaviors, sales channels and policy modes of specific equipment products, so that sales efficiency and market adaptability are improved, the tire removing machine is used as one of automobile maintenance equipment, the sales technical field covers the contents of customer demand identification, sales process decision support, supply and demand matching mechanisms, price dynamic adjustment, sales policy feedback optimization and the like, and the whole technical system mainly relies on elements such as historical sales data, market feedback information, equipment performance parameters, sales behavior paths and the like, and the sales strategies are constructed, evaluated and adjusted by combining mathematical models and algorithm reasoning tools, so that the management and support of the whole sales process are realized. The self-adaptive generation system of the tire changer sales strategy based on reinforcement learning is characterized in that a reinforcement learning method is utilized, an interaction mechanism between a sales environment and a decision agent is constructed, automatic generation and iterative optimization are carried out on the sales strategy, the targeted technical matters comprise sales scene modeling, sales behavior representation, state space construction, sales action decision, reward and punishment mechanism definition and the like, specifically, a strategy gradient algorithm is adopted to continuously adjust a sales strategy selection path in discrete time steps through setting sales state index parameters, and the strategy is corrected according to a set reward function in each round of interaction, so that target profit-oriented sales strategy output is formed, historical sales data is utilized as training samples based on the reinforcement learning framework, and the strategy is evaluated and updated through simulation environment, so that the self-adaptive generation of the sales strategy in different sales scenes is realized. In the prior art, the existing sales strategy generation mode focuses on the whole strategy path deduction, the sales state expression is mainly based on unified abstract indexes, the regional difference and the local supply and demand change are difficult to fully embody, the transformation fluctuation in multi-link interactive propulsion is mostly attributed to the result deviation instead of a specific operation node, when the customer feedback rhythm or service condition changes, the strategy response lacks process-level positioning support, the adjustment result is easy to stay on a macroscopic level, so that the local flow imbalance is accumulated, and the continuous sales performance and the strategy consistency are influenced. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides a self-adaptive generation system of a tire changer sales strategy based on reinforcement learning. In order to achieve the purpose, the invention adopts the following technical scheme that the tire changer sales strategy self-adaptive generation system based on reinforcement learning comprises: the supply and demand feature perception module is used for acquiring service node information of a sales area, dividing the sales area, distributing area node numbers, collecting transaction data, generating trend vectors, analyzing trend consistency, distinguishing supply and demand performance types and generating a supply and demand data feature area result; The conversion fluctuation identification module is used for obtaining the interactive operation conversion efficiency of each regional node according to the regional node number given to the supply and demand characteristic label in the supply and demand data characteristic regional result, extracting response time delay and protocol fluctuation, identifying the conversion efficiency deviation position, outputting the operation position number and generating a conversion fluctuation position result; the state operation association module extracts the client attribute, time delay, protocol change, feedback signal and lead time associated with the corresponding operation according to the operation position number in the conversion fluctuation position result, analyzes the relation with the interactive operation, establishes the corresponding description of the flow state and th