First, the decision-making level.
According to the role and responsibility of decision makers in the implementation of geological survey projects, decision-making can be divided into three levels: high-level decision-making, middle-level decision-making and grassroots decision-making.
(1) High-level decisions are strategic decisions. From the perspective of project establishment, high-level decision-making should stand at the height of national geological work, proceed from the needs of national social and economic development, and focus on solving the planning, direction and scale of geological survey and major geological problems involving national security; From the perspective of project implementation, high-level decision-making focuses on the overall operation of the project and the realization of the overall goal, focusing on solving problems such as organization, standards and systems.
(2) Middle-level decision-making is strategic decision-making. Middle-level decision-making is the development and extension of high-level decision-making Through middle-level decision-making, high-level decision-making is transformed into concrete plans for organization and implementation. This kind of decision-making includes making project plan, selecting project scheme, coordinating project implementation and synthesizing project results.
In decision-making, middle-level decision-making is very important, so we should pay attention to the downward deviation of decision-making. That is, it is not how to turn high-level decision-making into middle-level decision-making for concrete implementation, but the affairs of grass-roots decision-making as the content of middle-level decision-making.
(3) Grassroots decision-making belongs to executive decision-making, or business decision-making. Taking quality and cost as the center, be responsible for the specific business arrangement and organization in the implementation of work items.
According to the structure of decision-making, decision-making can be divided into procedural decision-making and non-procedural decision-making Procedural decision-making is a rule-based decision-making, which is usually repeatable.
Second, the decision criteria.
Scientific decision-making, in addition to scientific decision-making procedures and methods, must also meet the following five standards:
(1) There must be clear project objectives. For example, what geological problems are solved by the geological survey project, whether to submit the identified resource reserves or find out the geological conditions; For which department are these geological achievements provided and for what purpose; What is the expected result; Need to be clear when the project is established.
(2) The decision must have a reliable basis. An unfounded decision cannot be a scientific one. Similarly, decisions without scientific basis cannot guarantee the scientificity and correctness of decisions. The most basic scientific basis for decision-making is the research, demonstration and evaluation of technical and economic feasibility.
(3) Feasible decision guarantee. Whether the decision is correct or not is of course important. In the absence of manpower, financial resources, material resources and environmental protection, the correct decision may not achieve the desired results.
(4) It must conform to the economic principle. The feasibility of technology is the basis of decision-making. It is only technically feasible, but it cannot be used as a basis for decision-making. It must also be economically viable. The so-called economic principle, first, the benefit principle, is that it must be able to produce social and economic effects consistent with investment; The second is to achieve the above effect with minimum investment.
(5) Must have certain adaptability. The internal and external environment of the project is constantly changing. In the ever-changing environment, no decision can be fixed. When the environment changes, decisions must be adjusted in time according to the needs of project implementation. For example, in the design of a mining area, it is expected that the decision of ore bodies will be east-west. Therefore, the exploration line is designed to strike north and south, and the drilling holes are arranged along the exploration line, and the first drilling hole is completed, but no expected ore body is found. The exposure of surface engineering shows that the occurrence of ore bodies is north-south and inclined to the north. At this time, we must make a comprehensive analysis based on the obtained data and adjust and determine the next deployment, instead of blindly following the design and construction.
Third, the decision-making method.
There are many methods of project decision-making, which can be roughly divided into three categories: the first category is deterministic decision-making methods, such as linear programming analysis and inventory management analysis; The second category is decision-making under uncertain methods, such as pessimism method, minimax method, coefficient method, equal possibility method, regret value method, etc. The third category is risk analysis methods, such as expected value criterion analysis and marginal probability criterion analysis. The above-mentioned decision-making methods have not been widely used in geological survey project management for two reasons: first, it is difficult to quantitatively express the economic and social benefits of geological survey projects; Second, it is deeply influenced by the traditional geological work management, and the project management is still in the primary stage. In addition, too much emphasis on the particularity of geological work weakens the universality of project management, and the lack of research on geological survey project management is one of the reasons why scientific decision-making methods are rarely used in geological survey project management.
1. linear programming analysis method
Linear programming analysis is a deterministic decision-making method, which studies how to make full use of a certain amount of known manpower, material resources and financial resources in order to maximize the amount of tasks completed or minimize the consumption of people, money and materials. Linear programming analysis can solve the linear programming problem by solving the solution (maximum or minimum) of the objective function under a set of linear constraints. Among them, the constraints can be human resources, equipment, materials, funds and other constraints; The objective function can be maximum profit or minimum cost; Decision variables can be workload, output, sales volume, transportation volume, etc. The general form of the mathematical model of linear programming is as follows:
Solve a set of decision variables xj and satisfy a set of linear constraints:
Geological survey project management
Maximize or minimize the objective function:
Geological survey project management
Generally speaking, when establishing a mathematical model of resource management with linear programming, there are four steps:
The first step is to define the decision-making objectives and analyze the constraints;
Secondly, a mathematical model including a set of linear constraint equations or inequalities and the expression of the optimal linear objective function is established.
The third step is to solve the model by using the graphic method and algebraic method of linear programming to find the optimal solution of decision variables;
The fourth step is post-optimization analysis.
2. Decision-making under uncertainty method
(1) Wold method
Wald's method, also known as the maximum-minimum method or the minimum-median method, is the first choice for conservative pessimists. Its principle is: first, find out the minimum value of each decision goal in various States, and then choose a maximum value from the minimum value of these decisions, and the corresponding decision is the optimal decision. Expressed as:
Geological survey project management
Where q is the income of aj decision in Si state.
According to the min-max rule
Geological survey project management
Where Q* is the best return when Si is minimum and aj is maximum.
Then the optimal decision
Geological survey project management
For example, the person in charge of a project B must decide whether to apply the new technology P to the project. There are two possibilities for the application prospect of this project: good (S 1) and bad (S2). See Table 4-2 for the consequences (income Q) of new technology investment (a 1) and non-investment (a2). Solve the problem with Wald method.
Table 4-2 P Income Statement of New Technology
First, find the minF(Si, aj) of j = 1 2:
Geological survey project management
Recalculate: q * = max [-3,0 0] = 0 = f (Si, a2)=Q*
So a*=a2
Therefore, the optimal decision is not to use the new technology p. Conservative decision makers would rather not earn 200 thousand yuan than risk losing 30 thousand yuan.
(2) Max-Max method
Max method is called the maximum method, and it is the first choice for optimists who like to take risks, so it is also called the optimistic method. Expressed as:
Geological survey project management
Where q is the income of aj decision in Si state.
According to the max-max law
Geological survey project management
Where Q* is the optimal return under the conditions of maximum Si and maximum aj.
Then the optimal decision a*=aj.
For the above example, the calculation process of this method is:
Geological survey project management
So a*=a 1
This result shows that for an enlightened decision-maker, in order to get a profit of 200,000 yuan, he would rather take the risk of losing 30,000 yuan and use the new technology P.
3. Decision tree method
In order to visualize the decision-making method, the calculation process is drawn into a tree structure of points and branches, which is called decision tree and is suitable for the visualization of any decision-making method. Among them, the conditional node is represented by δ, and the decision node is represented by square □
Now, taking the decision data listed in Table 4-3 as an example, the steps of making decisions by using decision trees are explained. Suppose the probabilities of states Q 1, Q2, Q3 and Q4 are 0.2, 0.3, 0.3 and 0.2, respectively.
Table 4-3 Loss Expectation
Step 1: Build a decision tree model from left to right.
Tree building process: starting from the decision point on the left, divide several branches according to the number of alternative schemes, and each branch is marked with the brief content of the scheme. Each branch reaches a node, which is indicated by a ○, and the serial number of the scheme is marked by a ○. Then, starting from the node, each scheme is divided into several branches according to the number of possible target States, and each branch is briefly marked with the content of the target state and the probability of occurrence. In this example, as shown in Figure 4- 14, there are three schemes, and each scheme has four states.
Figure 4- 14 decision tree
Step 2: Make a decision from right to left. First, calculate the expected value of each scheme, and then mark the calculation results on the top of the status node, compare the expected values of each scheme, select the best scheme, and write the expected values on the top of the decision node to indicate the selection results. At the same time, draw a double tangent on the branch representing other schemes, indicating that these schemes have been eliminated.