CN115829600A - Batch-zero integrated marketing volume price optimization method and system - Google Patents
Batch-zero integrated marketing volume price optimization method and system Download PDFInfo
- Publication number
- CN115829600A CN115829600A CN202111090169.7A CN202111090169A CN115829600A CN 115829600 A CN115829600 A CN 115829600A CN 202111090169 A CN202111090169 A CN 202111090169A CN 115829600 A CN115829600 A CN 115829600A
- Authority
- CN
- China
- Prior art keywords
- price
- model
- volume
- retail
- wholesale
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a zero-batch integrated marketing volume price optimization method and a zero-batch integrated marketing volume price optimization system, which belong to the technical field of data statistical analysis, and comprise the following steps: analyzing the market competition characteristics of a specific area, and selecting an applicable market competition model; researching relevant factors influencing the quantity and price of a specific company, and establishing a simultaneous equation set model to find quantitative relations between the quantity and the price and among the factors; establishing a prediction model for the relevant variables which obviously influence the measure and price in the simultaneous equation set model; and designing an objective function, determining constraint conditions, and solving the optimal quantity price by using mathematical programming. The invention also discloses a batch zero integrated marketing volume price optimization system. The batch zero integrated marketing volume optimization method and system provided by the invention can give quantitative price strategies and sales volume targets, can estimate the maximum gross profit level, and are suitable for popularization and application.
Description
Technical Field
The invention belongs to the technical field of data statistical analysis, and relates to a zero-batch integrated marketing volume price optimization method and system.
Background
Batch-zero integration is a reflection of integrating existing resources to extend upstream and downstream of an industrial chain, is integration and reconstruction of the industrial chain, and is a unique novel business model. In the research aspect of wholesale and retail integrated optimization, scholars at home and abroad do a lot of work.
In the prior art, a diesel sales strategy optimization model is established for the sign difficulty, the constraint condition of the model is a diesel sales volume-price model, the objective function is that the profit of provincial company diesel sales is maximum, the diesel sales volume is changed into an explained variable, the international oil price change, the diesel sales price change, the competitor sales price change and the diesel sales volume change in the previous days are used as the explained variables, and modeling is carried out on the basis of an Arima model. Compared with the invention, firstly, the considered explanatory variables are incomplete, and the sales of competitors are not related, and the macroscopic economic situation and the market environment of the province are not related; secondly, the accuracy of a used prediction model is insufficient, the prediction model used in the literature is an Arima model and can only predict according to the change rule of the variable, the Arima-X model is introduced on the basis of the Arima model, the change rule of the variable is considered, the influence of exogenous variables on the variable is increased, the accuracy of a prediction result can be improved, the guiding significance of the Arima-X model on actual business work is not strong, the overall sales volume of diesel oil is taken as an explained variable in the literature, the wholesale sales volume and the retail sales volume of the diesel oil are respectively taken as the explained variable, the wholesale sales volume and the retail sales volume of the diesel oil are taken as the optimized batch zero structure, and the operability is higher in the actual business work.
The board of Dong Shaoyu constructs a diesel oil price model based on daily data of price and sales of diesel oil (wholesale) in a certain province of China Petroleum and gas group company, wherein an explained variable (dependent variable) is the diesel oil sales, and an explained variable (independent variable) is related price variables such as the diesel oil price and international crude oil price (WTI price). Compared with the invention, firstly, the considered explanatory variables are incomplete, and the sales volume of competitors, the influence of social resources, the macroscopic economic situation of the province and the market environment are not involved. And secondly, the method does not have a pre-judging function, and only obtains the sales volume corresponding to the price according to the historical rule.
The influence of discount rate on sales volume, the forecasting management of retail business of a certain medium and petrochemical company, the forecasting of gasoline sales trend of northeast China regions of medium and medium petroleum, the optimization of a finished oil sales system of a northeast China oil gas sales company and the competition strategy of petroleum enterprises in China are respectively researched by high defense, chengle, liuxin, wangzhong authority and ancient element, and considered variables and established models of the optimization have certain one-sidedness which is less than that of a variable pool and an optimization model designed by the invention.
In the aspect of foreign related research, the price prediction or the crude oil research is more favored, akrom uses Arima and Arima-X algorithms to predict the supply of crude oil and condensate, mishra, dhamija and Jeng respectively use Arima models to predict the natural gas price based on time series, predict the European Union subsidy and determine the price determinants thereof, and Gal discusses the influence of the uncertainty of the natural gas fuel cost on the capacity investment and price in the competitive power market. Although a prediction model in the prior art can achieve a certain technical effect, at present, a technical scheme for optimizing a batch zero structure and a batch zero price difference does not exist, and a batch zero integrated marketing price optimization method and system are urgently needed.
Disclosure of Invention
The invention aims to solve the technical problems and provides a batch-zero integrated marketing volume optimization method and a batch-zero integrated marketing volume optimization system.
In order to achieve the purpose, the invention mainly provides the following technical scheme:
first, the embodiment of the invention provides a zero-batch integrated marketing volume optimization method,
the method comprises the following steps:
analyzing the market competition characteristics of a specific area, and selecting an applicable market competition model; researching relevant factors influencing the quantity and price of a specific company, and establishing a simultaneous equation set model to find quantitative relations between the quantity and the price and among the factors; establishing a prediction model for the relevant variables which obviously influence the measure and price in the simultaneous equation set model; and designing an objective function, determining constraint conditions, and solving the optimal quantity and price by using mathematical programming.
Specifically, the market competition model includes: an oligopolistic competition model.
Further, the oligopolistic model comprises: a gulono model, a burtland model, or a starkeberg model.
Specifically, the research is used for researching relevant factors influencing the volume and price of a specific company, and establishing a simultaneous equation set model to find quantitative relations between volume and price and between volume and price, wherein the quantitative relations between volume and price and the quantitative relations between volume and price of each factor comprise the following steps: judging wholesale and retail market volume price influence factors according to market characteristics, and judging and adjusting specific relevant variables according to factors such as economy, supply and demand, policies, market structures, enterprise status and the like of different provinces; establishing four volume price influence models; and introducing data to perform empirical operation.
Specifically, the simultaneous equations model is given by the following formula:
wherein Q is Wholesale Is the volume of wholesale, Q Retail sale Is the retail quantity, P Wholesale Is a wholesale price, P Retail sale Is a retail price, P Competitor Is the price of a competitor, Q Competitor Is the sales volume of a competitor, C is a constant, a is a macroscopic economy, b is a virtual variable, f1 is a function of the wholesale volume, f 2 Is a function of retail quantity, f 3 Is a function of wholesale price, f 4 Is a function of the retail price.
Specifically, the establishing of the prediction model for the relevant variables that significantly affect the volume price in the simultaneous equation set model includes: establishing a time series prediction model according to self-periodicity and seasonal fluctuation; establishing a first metering prediction model according to the international oil price change; and establishing a second metering prediction model according to the macroscopic economic variables and the international oil price.
Furthermore, the time series prediction model is used for predicting macroscopic economic variables, the first metering prediction model is used for predicting wholesale prices and retail prices of competitors, and the second metering prediction model is used for predicting wholesale quantities and retail quantities of competitors.
Specifically, the objective function is given by the following formula: gross profit = sales income-acquisition cost + subsidy.
Specifically, the constraint conditions include: the wholesale price, the retail price, the wholesale quantity and the retail quantity of the Chinese petroleum diesel follow the boundary conditions of the price relationship; the relative market share of diesel oil retail and wholesale is not lower than the same year synchronization level; scenario-based planned completion rates of diesel sales (wholesale, retail); all wholesale and retail quantity and price variables are non-negative values.
Secondly, an embodiment of the present invention further provides a system for optimizing zero-batch integrated marketing volume price, where the system includes: one or more processors; a memory for storing one or more programs; the processor is configured to execute program instructions stored in the memory that when executed perform the batch zero unitary marketing volume price optimization method described above.
Thirdly, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, wherein the computer program, when executed by one or more processors, implements the method for optimizing zero-batch integrated marketing volume price.
Compared with the prior art, the invention has the beneficial effects that:
(1) Exploring a set of method capable of comprehensively measuring and calculating the optimal batch zero-volume price combination, giving a quantified price strategy and a sales volume target by using the method, estimating the maximum gross profit level, and realizing scene measurement and calculation;
(2) Analyzing the factors of each province influencing the wholesale, retail sales and price of China petroleum, and providing help for each province company to grasp market characteristics and deal with market changes;
(3) And establishing a diesel batch zero integrated marketing volume price optimization model of 31 provinces, wherein the data frequency is monthly, and a basis is provided for dynamically formulating marketing strategies for general companies and companies of various provinces.
The technical scheme of the invention can provide a quantitative price strategy and a sales volume target, can estimate the maximum gross profit level, and is suitable for popularization and application.
Drawings
FIG. 1 is a schematic diagram of a batch zero integration modeling concept;
FIG. 2 is a diagram of a theoretical model of competition in a finished oil market;
FIG. 3 is a schematic of the Gono model equalization;
FIG. 4 is a schematic diagram of the major factors affecting the fuel price;
FIG. 5 is a schematic diagram of the basic idea of argument prediction;
FIG. 6 is a schematic diagram of a planning solution;
FIG. 7 is a project study flow diagram;
FIG. 8 is a diagram illustrating the effect of non-conventional factors on underlying data;
FIG. 9 is a graph comparing actual values of national average retail prices with simulated values;
FIG. 10 is a comparison graph of smoothing before and after processing;
FIG. 11 is a schematic representation of the correlation between retail quantity prices.
Detailed Description
Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
1 diesel batch zero integration pricing optimization model design
1.1 model Overall framework
The diesel batch zero integration integral framework can be divided into four steps: s11, analyzing market competition characteristics of a specific area, and selecting an applicable market competition model; s12, researching relevant factors influencing the quantity and price of a specific company, and searching quantitative relations between the quantity and the price and relations between the quantity and the price by using a simultaneous equation set model; s13, establishing a prediction model for macroscopic variables, competitor volume prices and the like which obviously influence the volume prices; and S14, designing an objective function, defining constraint conditions, and solving the optimal quantity and price through mathematical programming. The overall modeling concept of the study is shown in figure 1.
1.2 selection of market Competition models
The complexity of the market simulation of the finished oil lies in that the market characteristics are different: differences in wholesale and retail market characteristics, differences in market entity control capability, differences in supply capability and customer demand characteristics \8230 \ 8230;. Based on the method, the theoretical derivation of the model is carried out according to the thinking of classifying the situations from simple to complex, static to dynamic, and the like, so as to establish the basis for quantitative analysis.
According to the analysis of market competition conditions by western economics, the competition of the finished oil market is simply classified, see fig. 2, and whether the oil industry is a natural monopoly industry or not is still analyzed in consideration of double attributes (commodity attributes and strategic attributes) of the oil industry chain. Thus, from the perspective of whether there is competition, the market can be divided into monopolized and nonproprietary markets. At present, the market is not basically monopolized except in some areas with difficult supply or severe situation, because the problems of safety, resources and the like are mild and the competition of the end market of the product oil does not involve basic strategic problems. Considering the market competition, the market can be divided into a complete competitive market and an incomplete competitive market according to the strength of the competition. A fully competitive market refers to a market structure that is fully competitive without any impediments and interferences. In this market type, there are a large number of buyers and sellers, buyers and sellers are the recipients of prices, resources are free-flowing, and information has integrity. In fact, the theoretical derivation of a fully competitive model is difficult to use in the existing product oil market: firstly, the assumption of a complete competition model is strong, and the simulation is difficult in reality; secondly, the number of competitive main bodies in the finished oil market in China is not large, the price is still not completely marketized at present, and the method is not suitable for using a complete competitive model. Although the monopoly model and the complete competition model have low applicability to the current product oil market in China, the research deduces the monopoly model and the complete competition model for reference considering that the competition conditions of certain regions accord with relevant theories. In the current finished oil market of China, a more applicable model is mainly an oligopolistic model. The study focused on three major models used in mainstream economics, the gulo model, the burtland model, and the starkeberg model.
The main objective of this study is to maximize the overall (retail + wholesale) benefit (revenue or profit) of the measurement by studying the volume-price balance between the retail and wholesale markets. Under this idea, a theoretical simulation is performed on the market competition situation, and its basic assumption includes the following 4 aspects: there are two markets in common, retail and wholesale markets respectively; the marginal cost of the enterprise is unchanged as C i (ii) a The enterprise will have its own price P i And sales amount Q i As the primary decision variable; determination of market equilibrium volume price: enterprise profit pi i Maximization of (e), profit = profit (R) i ) -cost (C) i )。
1.2.1 complete monopoly market
The model of competition (completely monopolized market situation) is not considered, and mainly refers to the market organization with only one or a few manufacturers in the whole industry. Specifically, the conditions of monopolizing the market are mainly as follows: first, only one or a few manufacturers and sales products are on the market; second, the manufacturer does not have any close substitutes for the goods it produces and sells; third, it is extremely difficult or impossible for any other vendor to enter the industry. In such markets, excluding any competing factors, monopolies control the production and marketing of the entire industry. Therefore, monopolies can control and manipulate market prices. In this case, the measure-price relationship function (assumed to be linear) P for Enterprise i i ＝f(Q i )＝a+bQ i (ii) a Benefit function pi for Enterprise i i ＝TR(Q i )＝P i Q i -C i Q i ＝f(Q i )Q i -C i Q i (ii) a Without special constraints, the profit maximization condition of an enterprise is as follows:
the solution can be obtained by the method,
thus, the optimum price and cost for a business without regard to competitionRelevant factor (C) i ) External macro-economic variable a. Enterprise optimal sales and cost related factors (C) i ) External macro-economic variable a and dose-price influence coefficient b. Among them, cost factor (C) i ) Including oil costs, marketing costs, inventory costs, and the like.
1.2.2 complete competitive market
A large number of buyers and sellers exist in the market, and each consumer and enterprise has no control force on the market price and can only passively accept the market price. In this case, the enterprise can only achieve the purpose of maximizing profit by adjusting the sales volume and cost thereof, and the profit maximization conditions of the enterprise are as follows:
the equation obtained after calculation is: c i = P = MR. That is, in a fully competitive market, the decision price of a business is related to a cost factor.
1.2.3 market for oligopeptides
1) Market for short (supply ≤ demand)
A few factories control the product sales throughout the market. And selecting different models for analysis according to the resource supply and demand conditions.
When the market supply is biased (supply is less than or equal to demand) or the sales volume is taken as a main decision for a long time, a Gonio model is taken as a theoretical analysis, competitive enterprises in the market achieve the maximum profit by controlling the sales volume, and a Gonio model equilibrium diagram is shown in figure 3. Taking the double oligo model as an example:
assume that the linear demand function for the market is: p = a + bQ i ＝a+b(Q i +Q j ) (ii) a The profit functions for enterprises i and j are:
the maximum value condition of the enterprise i is to take Q i Is 0, i.e.:
likewise, the first order conditions for Enterprise j are:
solving the above results, two decision functions of enterprises can be obtained, and an equation set is formed:
and obtaining a market equilibrium situation after solving, wherein the result is as follows:
that is, the decision sales volume of the enterprise is related to the cost factor of the enterprise, the cost factor of a competitor, an external macroscopic variable, a market volume and price relation coefficient and other factors; market price is related to the cost of all businesses in the market, external macro variables, and the like. The Gonio model equilibrium diagram is shown in FIG. 3.
2) Market for short (supply > demand)
When the market supply is loose (supply > demand), or the long-term decision is mainly considered by price, the Bertland model is taken as a theoretical analysis, and competitive enterprises in the market achieve the maximum profit by controlling the price. Also for example, the double oligo model: when two enterprises in the market produce the same cost, enterprise price decisions are related to cost.
Assume that the marginal cost of business i in the market is lower than the marginal cost of business j, i.e., C i ＜C j . In this case, the only Portland equilibrium is：
Wherein epsilon is less than 0, which is an infinitesimal quantity, the equilibrium price is slightly lower than the marginal cost level of enterprise j, enterprise i obtains all market demands, and enterprise j does not produce. Enterprise price decisions are related to costs, and sales decisions are related to costs, macroscopic variables, and volume-price coefficients.
3) Market for short (Multi-stage game)
Consider the situation of an oligopolistic market in the case of a multi-phase game:
while gaming at the present time, businesses in the marketplace are not aware of the competitor's decisions; when considering the knowledge of the enterprise about historical factors, the enterprise should consider the existing behavior of the competitor in the decision making. Theoretically, the analysis was performed using the Starkelberg model.
Suppose enterprise i is the leader, being the leading enterprise. Enterprise j is the follower. The gaming process proceeds in two steps: step one, enterprise i can not reverse after selecting sales; second, enterprise j is Q i The amount of sales is selected for a given situation. Market prices and business profits are realized in the second step.
The dynamic model is solved according to the inverse induction method, and the basic assumption and the functional form are consistent with the Guno model. Consider the market race scenario for a multi-phase game:
the reaction function for business j may be expressed as:
the profit for business i is expressed as:
importing the reaction function of the enterprise j into the profit function of the enterprise i to obtain a new profit function:
enterprise i to Q i When the first-order partial derivative of the enterprise I is 0, obtaining the balanced sales volume of the enterprise i; the reaction function brought into the enterprise j can obtain the balanced sales volume of the enterprise j, and further obtain the market balanced price.
According to the conclusion, the enterprise price decision is related to cost factors and macroscopic economic variable factors; the sales decision is related to cost, macroscopic variables and measure and price coefficients. Because the competitive characteristics of the market are different, the market characteristics need to be roughly judged in the actual operation process, and the model variables are considered according to the market competitive condition; in the process of model interpretation, market characteristics are also considered for interpretation. The theoretical model can solve the solution in the equilibrium state, but in the actual situation, the possible influence factors in the non-equilibrium state are also considered, and the influence factors of different models are shown in table 1.
TABLE 1 theoretical model related influencing factors
1.3 design of volume-price correlation model
The theoretical model focuses on the situation of a single market, but in practice the finished oil market can be divided into wholesale market and retail market according to the difference of sales patterns, and the two markets are in closer connection.
Problems with analog operations on a single market: a plurality of endogenous variables exist in the model, and the parameter estimators obtained by using least square regression one by one are inconsistent; the interrelationship between variables cannot be modeled with simple linear regression, requiring the establishment of an overall system and parameter estimation.
The simultaneous equation model is to distinguish various economic variables according to economic theory and certain assumed conditions, and establish a set of equations to describe the simultaneous relationship among the economic variables. The model describes that the causal relationship among the variables is bidirectional, and can reflect the operation process of the economic system more comprehensively and truly. According to research objectives, diesel marketing is concerned with four elements: wholesale price, retail price, wholesale quantity, retail quantity, namely, taking the four variables as main dependent variables to be researched. Meanwhile, according to the theoretical analysis, the factors influencing the method mainly comprise the sales volume, price and cost of the method and competitors, and key factors such as environment and resources; meanwhile, as the finished oil market in China is in the stage of gradually changing from price control to marketization, the related plans, assessments and the like of companies are important influencing factors. Based on the above analysis, we summarized the factors affecting the four marketing elements, see in particular fig. 4.
The process of establishing the simultaneous equations set model comprises three steps: firstly, judging wholesale and retail market volume price influence factors according to market characteristics; and judging and adjusting specific relevant variables according to factors such as economy, supply and demand, policy, market structure, enterprise status and the like of different provinces. And secondly, establishing four volume price influence models. And thirdly, introducing data to perform empirical calculation.
1.4 prediction of influencing factors
Before the operation of the mathematical programming model is carried out, the relevant variables need to be predicted: one part is the prediction of macroscopic economic variables, and mainly aims to delay the statistical distribution of the current macroscopic economic variable data and fail to meet the prediction requirement in practical application. And (4) establishing a time series prediction model according to the periodic and seasonal fluctuations of the macroscopic economic variables. The second part is the price of the competitor, and the measurement prediction model is established according to the international oil price change by the wholesale price and the retail price of the competitor, which is shown in figure 5. And thirdly, the sales volume of a competitor, and a metering prediction model is established according to the wholesale volume and the retail volume of the competitor and the macroscopic economic variable and the international oil price.
1.5 design of optimization model
The research relates to the problem of maximization of the overall benefits of two markets, and the theoretical analysis of a single market cannot be solved; besides the analysis of market conditions, the model is also limited according to specific operation conditions and operation targets; and further optimizing on the basis of a market competition model, and measuring and calculating by using a mathematical programming method in operational research.
Firstly, according to theoretical analysis, a basic mathematical expression of a model is set:
MaxΠ(P wholesale ，P Retail sale ，Q Wholesale ，Q Retail sale )≤0
s.t.g
(P Wholesale ，P Retail sale ，Q Wholesale ，Q Retail sale )≤0
(s Wholesale ，s Retail sale ，s Overall ) Not less than the specified portion
(P，Q，C，s)＞0
Objective function of the model: provincial companies maximize the gross profit in batch and zero integration. The constraints of the model are as follows: 1) The wholesale price, the retail price, the wholesale quantity and the retail quantity of the Chinese petroleum diesel follow the boundary conditions of the price relationship; 2) The relative market share of the diesel is not lower than the same-period level of the last year; 3) The relative market share of diesel oil retail is not lower than the same year synchronization level; 4) The relative market share of diesel wholesale is not lower than the minimum level of the last 12 months; 5) Scenario-based planned completion rates of diesel sales (wholesale, retail); 6) All wholesale and retail quantity price variables are non-negative values.
Designing a solving idea of the model: 1) Setting Lagrange function values lambda for the three types of constraints respectively; 2) Forming a Lagrange equation with the target function, and establishing a series equation set when the first-order partial derivative is 0; 3) Solving an equation set to obtain an optimal value; note that the assurance equation has a solution; 4) Currently the calculation process can be implemented in software.
On the basis of solving an equation set, designing a mathematical plan, and carrying out balanced solving: according to the constraint conditions, in the point set whose space accords with the correlation constraint, the point which can make the objective function reach the maximum is found, and the schematic diagram of the planning solution is shown in fig. 6.
Specifically, in practical application, the objective function and the constraint condition are modified and optimized. The method mainly comprises the following three points: firstly, considering the overall profit assessment mechanism and the related financial system of the company, and adjusting the objective function on the basis of the theoretical model. And secondly, the influence of oil price expectation on the diesel oil volume price is considered, and virtual variables such as price adjusting variables and the like are added. Thirdly, the quantity is properly increased or decreased in the aspects of import and export, economic variables and the like according to the specific province conditions.
1.6 application scope of model
A batch zero integrated volume price optimization model is a macroscopic metering model which can measure and calculate optimal batch zero volume price combination and is established by taking provincial companies as basic research objects under the conditions of surplus diesel oil resources and relatively stable market.
This model is not suitable for solving the following scenarios and goals:
1) Short-term policy and adversary behavior: the market has serious policy changes or competitors are abnormally sold, such as the government in a certain area suddenly implements the vehicle driving restriction policy, assaults the environmental protection and checks, and the competitors sell the policies in a big way.
2) Apparent resource supply and short supply: stable benefits can be realized from wholesale and retail sales channels, and the wholesale and retail channels greatly exceed the sales plan schedule operation, such as 10 months and 11 months in 2017 of part province.
3) The problem of logistics cost optimization: the batch zero integration model is based on proposing provincial marketing pricing strategies from the market competition perspective, obtains the target gross profit and share of a sales end, and does not contain the economic consideration of (once) logistics cost.
2 study procedure
The research of the measure and price cooperation model comprises three parts, namely variable selection, data preparation, model establishment and effect analysis, wherein the three parts totally comprise 10 steps. The project study flow is shown in figure 7.
1) Variable initial selection
By discussing preliminary framing of predicted objects and interpretation of variable ranges, i.e., dependent and independent variables, the variables add up to six major classes. The dependent variable requirements are authoritative, timely and strict in caliber, the independent variable requirements have stable data sources and are convenient to understand, monthly data from 2013 in 1 month are collected, and the calibers of all the monthly data are guaranteed to be consistent.
2) Setting of virtual variables
Because the sales of the Chinese petroleum product oil are greatly influenced by factors such as national price adjustment, seasonality, policy and the like, the numerical value change at a special time is abnormal, and an unconventional variable is set to eliminate the influence of the abnormal value on modeling.
3) Data pre-processing
The quantity and price of the product oil fluctuate too much along with the change of market environment, and the numerical value changes about twice, so that the quantity and price type variables are smoothed, the purpose is to eliminate the random fluctuation in the data, and otherwise, the precision of a prediction model is influenced. In addition, if the numerical change of the dependent variable at a special time point is abnormal due to an emergency or other reasons, a corresponding virtual variable is set to eliminate the influence of the abnormal value on the modeling.
4) Selecting independent variables related to each dependent variable
The degree of correlation between each independent variable and the dependent variable is determined by correlation analysis, and the independent variable with high correlation can be used as the independent variable entering the model. The selection of independent variables tends to have a practical impact on marketing efforts and represent as comprehensive a category as possible.
5) Setting an objective function and constraints
According to actual business requirements and relevant policies of China general oil sales companies, an objective function of gross profit maximization and quantity-price relevant constraint conditions are established.
6) Establishing a volume price preliminary model
And (4) establishing a multiple regression model by using the independent variables and the dependent variables obtained in the third step, wherein all parameters and test indexes of the model should meet the significance test requirements of the regression model, and correlation logics should meet actual business logics.
7) Determining the landscape variables of the dependent variables
The lunar data of the dependent variable still belongs to the time series data, and the regression model cannot completely solve the problem related to the time series residual error, so the problem is to determine the scenic variables of 1 month or several months in the future according to the residual error change rule of the fitting value and the actual value of the regression preliminary model.
8) Prediction method for establishing independent variable
Since the frequency of the independent variables used in the present subject is monthly, and no special research institution studies or publishes the predicted values, the present subject will build a prediction model of the independent variables by studying the periodic, seasonal, and trend changes of these variables themselves.
9) Forming model
Checking whether the parameters of the measure and price model meet the statistical requirements; and (4) performing simulation prediction on the actual operation condition according to the optimized result of the quantity-price collaborative model, and checking the prediction precision.
10 Tracking refinement and effect analysis
And tracking and monitoring the running stability of the model by combining with actual data, and checking whether the optimization result of the model meets the statistical requirement. And (3) performing simulation prediction on the historical condition by integrating the independent variable prediction model and the simultaneous equation model, and evaluating the optimization effect of the model by combining actual data.
3 variable selection and data preparation
3.1 variable selection
3.1.1 setting of entity variables
In order to meet modeling requirements, the SAS software is named in a mode of combining pinyin and English letters according to the properties of different variables, and all Chinese variables are translated into English codes. The above mentions that this study considers six major classes of indicators: the names of all entity variables of seven major indexes including internal resources, competitors, social resources, macro variables, plan execution and competition strategies and dependent variables are shown in table 2.
Table 2 entity variable naming code table
3.1.2 virtual variable settings
The virtual variable refers to a discrete variable reflecting abnormal changes of the original variable at a characteristic time point by setting numerical values of '0' and '1'. In some cases, the independent variable and the historical rules of the dependent variable cannot explain the reason of the individual abnormal value of the dependent variable, and only the dependent variable is considered to be influenced or interfered by some external strong factors, some of the external factors correspond to definite events, and the influence of the external factors can be reflected in the model through the form of the virtual variable. The role of the virtual variables in the model mainly embodies two points: the damage of singular points to the internal rules of the data sequence can be reduced; and the method provides a basis for adjusting the dependent variable when similar special events occur in the future.
1) Variable of price adjustment
At present, the sale of oil products is closely related to national price adjustment, and the monthly price adjustment condition is named according to the difference of price adjustment direction, times and amplitude. The up-regulation is represented by 'A', the down-regulation is represented by 'B', the price regulation amplitude is represented by '1' from '0 yuan/ton to 100 yuan/ton', the price regulation amplitude is represented by '2' from '100 yuan/ton to 200 yuan/ton', the price regulation amplitude is represented by '3' from '200 yuan/ton to 300 yuan/ton, the price regulation amplitude is represented by' 4 'above' 300 yuan/ton, the up-regulation and the down-regulation exist in monthly degrees, the total price regulation amplitude is represented by 'M' with a negative value, the total amplitude is represented by 'P' with a positive value, and the variable names of all price regulation variables are obtained. All price-adjusted variable designations are shown in table 3.
TABLE 3 CAVALENCE-ADJUSTING VARIABLE NAME CODE TABLE
2) Unconventional variables
There are many factors affecting the sale of diesel oil by analyzing the historical data of the sale of product oil, as shown in fig. 8, and periodic changes of seasons, sudden changes of external environment, etc. all have an effect on the sale of diesel oil.
The study also incorporates the above mentioned types of influencing factors into the model by designing unconventional variables, further modifying the results of the model. The unconventional variables include: seasonal variables, holiday variables, resource variables, policy variables, etc., see table 4.
TABLE 4 unconventional variable naming table
3) Variation of the sense of the scene
As can be seen by the comparison of actual versus simulated values for retail sales in FIG. 9, the actual retail diesel volume curve is higher than the simulated curve over time and lower than the predicted curve over time. The phenomenon shows that the residual error is relevant, and the retail quantity of the diesel oil is influenced by various economic factors and is also governed by the change rule of the retail quantity. That is, the periodic fluctuations in the retail quantity of diesel fuel are not consistent with the periodic fluctuations in the macroscopic economy, which has either a lead or lag presence.
After the scene virtual variable is introduced, the influence related to the residual error can be eliminated, so that the predicted value is guaranteed to be the average value of the actual value. And the value of the virtual variable of the scenery is determined according to the sign of the residual value, if the residual value is less than zero, the value of the virtual variable of the scenery is defined as 0, otherwise, the value of the virtual variable of the scenery is defined as 1.
After the introduction of the landscape gas variables in the study, the d.w value of the model can be adjusted up from about 0.3 to about 1.5, which is greatly improved as detailed in table 5. Value of D.W is the degree of fitting R divided by 2 In addition, another index of the model was evaluated. The d.w value is mainly used to check whether the random error term has sequence-related problems in the form of first-order autoregressive, and is suitably around 2.0.
TABLE 5 Effect of scene variables on the model
3.2 data preparation
Since it may not be good to use part of the data directly for modeling, it needs to be preprocessed. There are two cases of data preprocessing in this project. The first is to smooth the variable itself to obtain a moving average value; and secondly, analyzing the integral panel data, identifying the inflection point of the market rule, and selecting proper modeling data.
1) Smoothing process
The sales savings have larger fluctuation in a given time period, and the sales can be different by more than 3 times due to different external environments in light and busy seasons, so that the sales are smoothed, and the influence of the extreme value on the model is weakened. The number of variables to be smoothed in each province model is 9, and when all the smoothed variables are named, the original variable name is added with a letter "0", for example, the name after "pcry" logarithmic transformation is "pcry0". The wholesale quantity is taken as an example for demonstration, and the change trend before and after the smoothing treatment is shown in figure 10.
2) Selecting suitable modeling data
Taking Shanxi province as an example, the basic data of the Shanxi province is shown in FIG. 11, and the correlation among retail quantity prices is analyzed. From 2013, the quantity and price are positively correlated, and the correlation coefficient reaches 0.46; beginning in 2015, a negative correlation began to appear between volume prices. Therefore, data after 2015 was selected for modeling.
The embodiment of the invention also provides a zero-batch integrated marketing volume optimization system, which comprises the following steps:
the method comprises the following steps: one or more processors; a memory for storing one or more programs; the processor is configured to execute program instructions stored in the memory that when executed perform the batch zero unitary marketing volume price optimization method described above.
The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by one or more processors, implements the above-mentioned batch zero-volume marketing volume optimization method.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose systems may also be used with the teachings herein. The required structure for constructing such a system will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In addition, the memory may include volatile memory in a computer readable medium, random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a system for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction system which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. The use of the phrase "including an" as used herein does not exclude the presence of other, identical elements, components, methods, articles, or apparatus that may include the same, unless expressly stated otherwise.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention. Furthermore, although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.
Claims (11)
1. A zero-batch integrated marketing volume price optimization method is characterized by comprising the following steps:
the method comprises the following steps:
analyzing the market competition characteristics of a specific area, and selecting an applicable market competition model; researching relevant factors influencing the quantity and price of a specific company, and establishing a simultaneous equation set model to find the quantitative relation between the quantity and price and the factors and the relation between the quantity and price; establishing a prediction model for the relevant variables which significantly influence the volume price in the simultaneous equation set model; and designing an objective function, determining constraint conditions, and solving the optimal quantity and price by using mathematical programming.
2. The method of claim 1, wherein the market competition model comprises: an oligopolistic competition model.
3. The method of claim 2, wherein the oligoprobe competition model comprises: a gulono model, a burtland model, or a starkeberg model.
4. The method of claim 1, wherein the step of studying the relevant factors affecting the volume and price of a specific company, and establishing a simultaneous equation set model to find the quantitative relationship between the volume, price and each factor and the relationship between the volume and price comprises: judging wholesale and retail market volume price influence factors according to market characteristics, and judging and adjusting specific relevant variables according to economic, supply and demand, policy, market structure and enterprise status factors of different provinces; establishing four volume price influence models; and introducing data to perform empirical operation.
5. The method of claim 1, wherein the simultaneous equations model is given by the formula:
wherein Q is Wholesale Is the volume of wholesale, Q Retail sale Is the retail quantity, P Wholesale Is a wholesale price, P Retail sale Is a retail price, P Competitor Is the price of a competitor, Q Competitor Is the sales volume of a competitor, C is a constant, a is a macroscopic economy, b is a virtual variable, f 1 Is a function of the wholesale quantity, f 2 Is a function of retail quantity, f 3 Is a function of wholesale price, f 4 Is a function of the retail price.
6. The method of claim 1, wherein the establishing a predictive model of the variables associated with significant impact on the measure price in the simultaneous equations set model comprises: establishing a time series prediction model according to self-periodicity and seasonal fluctuation; establishing a first metering prediction model according to the international oil price change; and establishing a second metering prediction model according to the macroscopic economic variables and the international oil price.
7. The method of claim 6, wherein the time series prediction model is used to predict macro-economic variables, wherein a first metric prediction model is used to predict wholesale prices and retail prices of a competitor, and wherein a second metric prediction model is used to predict wholesale quantities and retail quantities of a competitor.
8. The method of claim 1, wherein the objective function is given by the formula: gross profit = sales income-acquisition cost + subsidy.
9. The method of claim 1, wherein the constraints comprise: the wholesale price, the retail price, the wholesale quantity and the retail quantity of the Chinese petroleum diesel follow the boundary conditions of the price relationship; the relative market share of diesel oil retail and wholesale is not lower than the same year; the completion rate of the diesel sales volume planning of each scene is divided; all wholesale and retail quantity and price variables are non-negative values.
10. A zero-batch integrated marketing volume optimization system is characterized by comprising:
the method comprises the following steps: one or more processors; a memory for storing one or more programs; the processor is configured to execute program instructions stored in the memory that when executed perform the batch zero unitary marketing volume price optimization method of any of claims 1 to 9.
11. A computer-readable storage medium, having a computer program stored thereon, wherein the computer program, when executed by one or more processors, implements the batch zero unitary marketing volume price optimization method of any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111090169.7A CN115829600A (en) | 2021-09-16 | 2021-09-16 | Batch-zero integrated marketing volume price optimization method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111090169.7A CN115829600A (en) | 2021-09-16 | 2021-09-16 | Batch-zero integrated marketing volume price optimization method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115829600A true CN115829600A (en) | 2023-03-21 |
Family
ID=85515151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111090169.7A Pending CN115829600A (en) | 2021-09-16 | 2021-09-16 | Batch-zero integrated marketing volume price optimization method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115829600A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117172534A (en) * | 2023-08-17 | 2023-12-05 | 中石油云南石化有限公司 | Optimized control method for avoiding market risk through balance pricing of finished oil and crude oil |
-
2021
- 2021-09-16 CN CN202111090169.7A patent/CN115829600A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117172534A (en) * | 2023-08-17 | 2023-12-05 | 中石油云南石化有限公司 | Optimized control method for avoiding market risk through balance pricing of finished oil and crude oil |
| CN117172534B (en) * | 2023-08-17 | 2024-05-14 | 中石油云南石化有限公司 | Optimized control method for avoiding market risk through balance pricing of finished oil and crude oil |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Rao et al. | Compound mechanism design of supplier selection based on multi-attribute auction and risk management of supply chain | |
| Kellner et al. | An a posteriori decision support methodology for solving the multi-criteria supplier selection problem | |
| Liu et al. | Supply chain outsourcing under exchange rate risk and competition | |
| Marchi et al. | A joint economic lot size model with financial collaboration and uncertain investment opportunity | |
| Laitner et al. | Technological change and the stock market | |
| Boehm et al. | Convex supply curves | |
| Weidlich | Engineering interrelated electricity markets: an agent-based computational approach | |
| Klein | An essay on the accuracy of economic prediction | |
| CN115829600A (en) | Batch-zero integrated marketing volume price optimization method and system | |
| Broer | Neoclassical theory and empirical models of aggregate firm behaviour | |
| US20210056625A1 (en) | Artificial intelligence and machine learning in a clustering processes to develop a utility model for asset allocation and automated rebalancing of exchange traded funds | |
| Kuvalin et al. | The impact of price dynamics in Russia on monetary policy and economic development | |
| US20170004521A1 (en) | Systems and methods for generating industry outlook scores | |
| CN116050987A (en) | Prediction method and device for material replenishment, storage medium and processor | |
| CN115564265A (en) | Power grid enterprise digital transformation evaluation method based on evaluation model | |
| CN114692973A (en) | Data processing method and device | |
| WO2018005708A1 (en) | Systems and methods for generating industry outlook scores | |
| Schembri | Environmentally adjusted domestic product and emission control policies; a dynamic simulation modelling approach: an example from the French carbon dioxide emission prospects | |
| Ma et al. | Order acceptance policy for make-to-order supply chain | |
| Tikhonov | Adaptive production planning for enterprises during an economic crisis | |
| Jamshidnejad | Project Portfolio Selection Based on Risk Index | |
| Pakhomov et al. | The use of econometric tools for cost management analysis | |
| Ross et al. | Modeling of investment in it-business | |
| Ross et al. | Financial costs optimization for maintaining critical information infrastructures | |
| JP7495209B2 (en) | Management index evaluation device, management index evaluation method, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |