TWI225189B - Prediction system - Google Patents

Abstract

A prediction system comprises a prediction model forming unit 1 for predicting the subject of prediction, including a prediction model having a plurality of parameters; a parameter value storage unit 2 for storing a plurality of parameter values; and an input-and-output data storing means for storing time-series input and output data. A static parameter identifying means determines a restrictive condition to be imposed on parameters, on the basis of stationary input and output data. A dynamic parameter identifying means determines the values of the parameters within a range meeting the restrictive condition determined by the static parameter identifying means on the basis of nonstationary input and output data stored in the input-and-output data storage means, and sends the values of the parameters to the parameter value storage means.

Description

1225189 玖 发明, description of the invention [Technical field to which the invention belongs] The present invention relates to a prediction system, which includes a prediction model for online prediction of natural phenomena and man-made phenomena. For example, the prediction of the rate of rainwater flowing into a sewage plant and the quality Forecast, river water level forecast, dam water storage forecast, and water, natural gas and electricity demand forecast. [Previous technology] Forecasting technology is one of the important basic technologies mainly used in the field of utilities. It is used to predict the demand for water and the flow rate of rainwater into sewage plants. Predictive control, and economic sectors used to predict stock prices. In general, forecasts are obtained by processing the data available at the start of some processing (hereafter referred to as "current time"). Typical prediction methods use some prediction models to make predictions. Prediction methods are roughly divided into the use of prediction models including numerical equations (white box approximation) based on laws of physics and chemistry, and the use of statistical processing based on real data, such as neural network learning, etc. , Or a prediction model constructed by system identification (black box approximation). The advantage of the former prediction method is that because of using the laws of physics (or chemistry), it is possible to make predictions that conform to (not conflict with) the laws of physics (or chemistry), and because in most cases the parameters have physical and chemical meanings So, you can adjust the model parameters quite easily. On the other hand, the disadvantage of the previous prediction method is that it is difficult to determine a parameter determination guideline when there are many parameters to be determined, that is, a priority guideline for determining parameters for adjustment. As a result, accurate predictions cannot be made. Another disadvantage of the previous prediction method is that the variables representing the relationship between cause and effect are neatly divided into cause variables (inputs) and result variables (outputs), and because in most cases no previous information about the result variables is used Therefore, the prediction accuracy will be reduced. The latter prediction method is to construct a model based on a predetermined deductive method or identification deductive method so that the model conforms to real data. Therefore, the model is at least suitable for accurate identification or learning of data, and can make accurate predictions even when the data is not used for most of the learning or identification. On the other hand, the disadvantage of the latter method is that when it is desired to change (adjust) the parameters to improve the prediction accuracy of the specific input data corresponding to the specific input data, it is impossible to adjust or fine-tune the decision through learning or recognition based on a predetermined deductive method. Parameters, because the parameters do not have physical and chemical significance and the guidelines for adjusting the parameters are unknown, readjustment or fine-tuning is not possible, and the predicted results will contradict the physical (or chemical) laws that must be met in accordance with the laws of nature. law. Prediction that violates the laws of physics such as conservation of mass or conservation of energy is one of the factors that predicts the black box approximation. The conventional prediction technology recognizes the advantages and disadvantages of the white box approximation method and the black box approximation method and selectively uses the white box approximation method and the black box approximation method according to the situation. The inventor of the present invention has a Japanese patent application number of 1 3 2 9 9 8 / 2 0 0 1 Parameter adjustment of "Apparatus for Adjusting Parameters of Process Model, and Apparatus and Method of Assisting Adjustment" (3) (3) 1225189 The entire method proposes a guide on parameter adjustment that addresses one of the shortcomings of the white box approximation method. The problem caused by the unscrupulous. This previously proposed method of parameter adjustment is based on the analysis of the sensitivity of the parameters to adjust the parameters. Unsatisfactory prediction accuracy due to unused past information related to outcome variables can be improved by correcting predictions based on past data related to outcome variables. For example, when the prediction mode is described by a differential equation: χ = / (χ, θ, υ) .......... (1) y = 1ι (χ, θ, ν) .... ...... (2) By setting appropriate observers, predictions can be corrected, and predictions can be made by using past information about the result variable (y). In the expressions (1) and (2), X, Θ, u, y, f, and h are the state variables of the model, parameters, input (cause variable), output (result variable), and (non-linear) vector. And (non-linear) output functions. When constructing the pre-model by white-box approximation, we have tried to overcome the shortcomings of the white-box approximation by the methods described above. Many situations require the prediction model to be constructed by the black-box approximation. In most cases, it is difficult to apply the white box approximation to the specific forecast of the (forecast) system. In the system, for example, the outcome variables that need to be predicted, such as stage, flow, and water volume, and the reason variables such as rainfall are To a certain extent, it is very unsophisticated ', but internal structures such as rainwater outflow treatment are quite unsophisticated. In this case, it is preferable to construct a prediction model by a black box approximation. In particular, when considering a large amount of time series data stored in a data server or the like, in order to effectively use the data, the prediction using a black box model is more certain. From this point of view, the inventor of the present invention in Japanese Patent Application No. 7- (4) (4) 1225189 please apply for 61866/1990 "Apparatus for Prediction Amount of Inflowing Rainwater, and Method of Predicting Amount of Inflowing Rainwater", Japanese Patent Application No. 224 7 6 6/1998 `` Apparatus for Prediction Amount of Inflowing Rainwater, and Method of Predicting Amount of Inflowing Rainwater '', and Japanese Patent Application No. 2423 1/2002 `` Apparatus for Predicting River Stage '' proposed some use of the black box model method of prediction. These prediction methods have feature points and can actually provide highly accurate predictions. However, these methods are not free from the above disadvantages of the black box approximation. Therefore, when these methods cannot obtain accurate predictions or when they need to improve the prediction accuracy of specific output data related to specific input data, the guidelines for reading and adjusting the identified parameters are not good, And it is possible to predict that it does not conform to the laws of physics (chemistry) represented by the law of conservation. [Summary of the Invention] The present invention has been made in consideration of the above circumstances. Therefore, an object of the present invention is to provide a prediction system when a system identification method using a discrete-time system model, which is typically a black box approximation method, is provided. The white box approximation considers a prediction model of the laws of physics (the law of conservation). According to the present invention, a prediction system for predicting a prediction target includes a prediction model forming unit for predicting the prediction target, including a prediction model having a plurality of parameters; a parameter 値 storage unit for storing a plurality of parameters 値; an input and Output data storage mechanism for storing time-series input and output data for prediction. (5) (5) 1225189 Input and output data; static parameter identification mechanism, which assumes a static response (output) to a fixed input, or The static output data of the fixed input data determines the restriction conditions to be imposed on the parameters; and the dynamic parameter identification mechanism determines the compliance with the static parameter identification based on the non-static input data and output data stored in the input and output data storage mechanism. The parameter 内 is within the limits determined by the institution, and the parameter 値 is transmitted to the parameter 値 storage institution. Unlike the conventional method of directly identifying dynamic parameters, the present invention uses a static parameter identification mechanism that takes into account the physical constraints of the steady state, puts constraints on the parameters, and can predict so that the steady state prediction is accurate The degree will not decrease. Since the relationship between the steady-state input and output corresponds to the law of physical conservation or in most cases in prediction systems, it can constitute a prediction model that considers the laws of physics indirectly. Therefore, even in an unstable state, it is possible to avoid making predictions that conflict with the law of conservation, and to ensure that a physically acceptable prediction is obtained. In the prediction system according to the present invention, the prediction model forming unit includes a coefficient parameter section (automatic regression section) of a prediction model that works linearly on time series output data and a coefficient parameter section (moving average) that works linearly on time series input data ), And the static parameter identification mechanism, which applies constraints to the parameter 値 so that the ratio bb / aa is fixed, where aa is the sum of the coefficients 自动 of the automatic regression section, and bb is the coefficient of the moving average section The sum of 値. According to the present invention, the necessary conditions to be satisfied by the prediction model conforming to the conservation of quality can be expressed by a very simple -9- (6) 1225189 single exponent (restriction condition) by using a discrete-time linear transfer function model. This restriction condition is a parameter The ratio between 値 is fixed. The discrete-time linear transfer function modules ARX model, ARMAX model, and BJ model are compared with the prediction model of the ink tank model. Therefore, by adjusting the constraint condition that the ratio between the sum of the parameters 固定 is fixed, it is possible to construct a physical prediction in a similar way. When needed, one of the adjusted parameters of the prediction model can be easily changed according to simple ′. In the prediction system according to the present invention, the static parameter recognition machine imposes restrictions such that the total number a of the parameter 自动 of the automatic regression unit is usually a fixed positive 値 which is close to zero and expressed as aa = ε, 0 < ε < < 1. According to the present invention, the parameters can be adjusted so that the sum of the parameters 自动 of the automatic return and the sum of the parameters of the moving average section remains fixed, and a very promising adjustment can be achieved. In addition, predictions can be made even if fixed or very slowly changing substances act on the prediction system. The accuracy of the prediction can be limited to the smallest unavoidable degree. The satisfactory prediction result can be given to the biased prediction system to make the prediction When the input to the system is zero, the output does not fall to zero. In the prediction system according to the present invention, the dynamic parameter decision machine restricts the conditions to identify the parameters 値, and the restriction conditions require that the parameters of the automatic regression section and the moving average section previously determined by the static identification mechanism are respectively selected by using the optimal Mechanism and fixed using non-stationary input and output data stored in the input and material storage mechanism According to the present invention, first determine the conditions that should be met in any situation, rather than the dynamic parameters used by the conventional black box approximation The straight aggregate type includes the determination of whether a is non-returning to the index structure that satisfies the black box near, and 'so that it is unified. It is very easy to construct a limitation based on the state parameter 値 total output capital 〇 Recognition -10- (7) (7) 1225189, the restriction conditions will be stored and the parameters will be identified to meet the dynamic data of input and output. Therefore, it is possible to prevent the accuracy of recognition from being lowered. As a result, the identification can be obtained even under the condition that the general identification method will cause the number of parameters to diverge, thereby improving the robustness of the prediction. For example, when the conventional adaptive identification method represented by sequential least squares is used to sequentially improve the accuracy of predictions, the parameters to be identified may diverge and if the identification data contains abnormal data, no order can be made. Satisfactory forecast. When the method according to the present invention is used for adaptive prediction, since a limitation is imposed on the parameter 値 total, this phenomenon can be suppressed, and thus a prediction device having strong tolerance to abnormal data can be constructed. Although sufficient data cannot be obtained in most cases when predicting natural phenomena such as river water levels and sewage flow rates, these situations can still constitute a certain degree of reliable prediction device. According to the present invention, a prediction system for predicting a prediction target includes: a prediction model forming unit for predicting the prediction target, including a prediction model having a plurality of parameters; a parameter 参数 storage unit for storing a plurality of parameters 値; input and output A data storage mechanism for storing time series input and output data collected at a predetermined prediction period; a first dynamic parameter identification mechanism for determining parameters based on non-fixed input and output data stored in the input and output data storage mechanism値; static parameter identification mechanism for determining the constraints to be imposed on parameters based on static input and output data; and a second dynamic parameter identification mechanism, based on non-static input and output stored in the input and output data storage mechanism Data, adjust the parameter 値 determined by the first dynamic parameter identification mechanism so that the parameter 値 is within the limits of the static parameter identification • 11-(8) (8) 1225189, and the parameter 値_Send to parameter 値 storage mechanism. According to the present invention, the use of the second dynamic parameter identification mechanism improves the prediction accuracy of a particular part of the output and enables more precise prediction. The prediction system further includes a display device for simultaneously displaying the input and output data stored in the input and output data storage mechanism and the prediction data corresponding to the input and output data and provided by the prediction model forming unit; and a marking device for The part of the prediction model required to reduce the prediction error and the cause variable that is considered to cause the error are marked; among them, the second dynamic parameter identification mechanism identifies the parameter of the cause variable marked by the labeling device to reduce the Some prediction errors. According to the present invention, the difference between the prediction 値 and the real output can be visually recognized, and a precise prediction can be obtained, and the degree of improvement in the prediction accuracy of the part whose prediction accuracy is to be improved can be visually recognized. The prediction system further includes a display device for simultaneously displaying the input and output data stored in the input and output data storage mechanism and the corresponding input and output data and the prediction data provided by the prediction model forming unit; and adjustable parameters Marking and display device, which is used to mark the prediction model and at least two parameters for selecting the cause variables for manual tuning. For the part of the prediction model, the data displayed relative to the display device needs to be reduced. Predictive errors, and causal variables that are considered to be the cause of the errors. According to the present invention, the prediction accuracy of the part that needs to reduce the prediction error can be manually adjusted, and the degree of difference between the prediction 値 and the real output can be visually recognized. In particular, after the -12- (9) (9) 1225189 parameter is selected by the adjustable parameter selection device, only one parameter needs to be changed, so the adjustment can be easily obtained and the prediction result due to the adjustment is in Visually recognizable. [Embodiment] First Embodiment A prediction system according to a first embodiment of the present invention will be described with reference to the drawings. Fig. 1 shows the basic configuration of the prediction system according to the present invention applied to the prediction of the inflow amount into the sewage field. It is mentioned that the inflow prediction is a specific embodiment of the operation of the prediction system and the contents described below can be applied to any prediction system. Refer to Figure 1 'The forecasting system applied to the sewage system inflow system 1 includes rainwater measurement device 1 1, inflow measurement device 12 and rainwater inflow processing 1 3' rainwater inflow processing 丨 3 represents the rainfall observation field and inflow Measure the passage of outflows and inflows between observation fields. The rainfall measuring device 11 continuously or at predetermined intervals measures the rainfall amount at a plurality of inflow locations in the sewage field included in the sewage system inflow system 1. The inflow measuring device 12 measures the inflow into the sewage field included in the sewage field inflow system 1 continuously or at a predetermined period. As shown in FIG. 1, the prediction system of the present invention includes a prediction model device 2 and an input and output data storage mechanism 3. The prediction model device 2 includes a prediction model formation unit 21 and a parameter / storage storage unit 22, and the prediction model is formed -13- (10) (10) 1225189 Unit 21 includes a prediction model, an input device for inputting data into the prediction model, and an output device for providing the data provided by the prediction model. The parameter / storage unit 22 is used to store the prediction model. The parameters 値 of the prediction model of the forming unit 21, and the input and output data storage mechanism 3 are used to store the rainfall data provided by collecting the rainfall measurement device 11 at a predetermined cycle and the inflow measurement device 12 Time-series data obtained from the inflow data. The input and output data storage mechanism 3 is connected to the static parameter identification mechanism 4 'by using the data stored in the input and output data storage mechanism 3 and checking the characteristics of the sewage field inflow system 1 to determine the device to be applied to the prediction model device. The parameter "limitation condition of parameter" in the storage unit 22 'and the parameter number determination mechanism 5 connected to the identified parameter number 5 are used to determine the number of parameters of the prediction model forming unit 22. The dynamic parameter identification mechanism 6 is connected to the input and output data storage mechanism 3 ° The dynamic parameter identification mechanism 6 is connected to the input and output data storage mechanism 3 ° The dynamic parameter identification mechanism 6 determines the parameters of the prediction model forming unit 22 to be used by storing The rainfall data and inflow data in the input and output data storage mechanism 3 accurately determine the dynamic relationship between the rainfall data and the inflow data provided by the inflow system 1 of the sewage plant. The input and output data storage mechanism 3 includes an identification / confirmation data storage mechanism 31 and a prediction data storage mechanism 32. Identification / certification data storage unit 31 will identify the parameters used to construct the prediction model and confirm the accuracy of the prediction. The prediction data storage mechanism 32 stores the passing data obtained in the passing period between the predicted point required for the real online prediction and the predetermined passing point. -14- (11) (11) 1225189. The operation of the first embodiment will be explained. The rainfall measurement device 11 and the inflow measurement device 12 respectively measure rainfall data and inflow data in a predetermined period of time and store them in the input and output data storage mechanism 3 as time series data. Since no prediction model was constructed at the beginning of *, at the beginning, rainfall data and inflow data will be stored in the identification / confirmation data storage mechanism 31. Next, the static parameter identification mechanism 4 uses the time-series data stored in the identification / confirmation data storage mechanism 31 to calculate a restriction condition. A significant difference between the present invention and the conventional prediction model construction method based on system recognition lies in the use of a static parameter identification mechanism 4, which is the most important component of the present invention. The operation of the static parameter identification mechanism will be specifically explained. The effective rainfall and outflow coefficients required to determine the constraints to be imposed on the parameters of the prediction model for predicting the inflow will be explained. The effective rainfall is the rainfall that actually flows into the sewage field, which is equal to the total rainfall minus the rainfall infiltration into the surface. The ratio of effective total rainfall to total rainfall will be called the outflow coefficient. Assuming the outflow coefficient k is fixed, the relationship between rainfall R (t) and inflow Q (t) is expressed by equation (3) as t-tf tt.f |; Q (t) -k ^ R ( t) (3) t «0 t« o where t is a parameter representing time and tf is the time when the rain stops. Since rainfall is regarded as the sum of effective rainfall (inflow) and rainfall infiltration into the surface, equation (3) is regarded as an expression representing the law of conservation of water. Therefore, the ability to obtain correct and effective rainfall prediction is obviously -15- (12) 1225189 is a necessary condition for an acceptable prediction model. The inflow amount prediction made by the prediction model forming unit 21 including the transfer function model of the discrete-time system will be explained. The relationship between rainfall r (t) and inflow Q (t) is expressed by expressions (4), (5), and (6). Automatic regression part A (q'1) = l + aiq-1 + a2q " 2 ^ ... + anq-n… (5) Moving average part B (q-1) == biqel + b2q * 2 + .. . + binq ", n… (6) Among them, it is a shift operator that indicates that the time is one step behind. Assuming R (t) Q (t) R0 (quantitative), that is, continuous fixed rainfall, the inflow Qo (quantitative) is expressed by equation (7) £ \ + b2 + Q 「_SR 〔+ a! + a2 + ". + an -R〇 (7) If the constant rainfall continues, then we know from formula (3) that the fixed rainfall Qo is expressed by formula (8). Q〇 = k. R〇 ·. · (θ ) Therefore, h + b2 +. "+ Bm (9) l + ax + a2 + --- + an Equation (9) represents the limiting condition: the sum of the parameters 移动 of the moving average part relative to the parameters of the automatic regression part 値The combined ratio must match the outflow coefficient k. Since the outflow coefficient k can be checked by using the data stored in the identification / confirmation data storage mechanism 31, it is possible to determine the sum of the parameters 値 of the moving average section -16- (13) (13) 1225189 relative to The ratio of the sum of the parameters of the automatic regression section is appropriate. In this way, the static parameter identification mechanism 4 determines the restriction conditions to be imposed on the parameters, and the dynamic parameter identification mechanism 6 determines the parameters 値 in a range that meets the limitation conditions. In general, although the restriction conditions are not convenient for the identification of the dynamic parameter 値, as long as the restriction conditions are met, the outflow coefficient can be regarded as the indicator of the conservation of water quantity. Therefore, using any parameter determined under this restriction, the prediction 値 predicted by the optional prediction device will not conflict with the law of conservation of mass. Therefore, although the transfer function of the discrete-time system, that is, the model, looks similar to a black box, the constraints can still indicate the quality. Although it is assumed here that the outflow coefficient is fixed, in some cases, outflow The quantity coefficient depends on the rainfall intensity and rainfall in the true inflow forecast. In this case, Equation (9) cannot be satisfied and it is learned that using a linear transfer function model is not appropriate. In this case, by using a non-linear model to set different fixed chirps for different rainfall intensities, constraints on these parameters can be determined. For example, the non-linear model can be Japanese Patent Application No. G 1 866 / 1 9 9 9 、 Rainwater Inflow Predicting Apparatus and

The Hammer stein model described in "Rainwater Inflow Predicting Method". For example, if the outflow coefficient can be approximated by a quadratic polynomial, the intensity of the Hammerst e in model with a quadratic polynomial can be considered at the input値 for the static parameter identification mechanism 4. The conditions used in this case and the parameters of the moving average section • 17 · (14) (14) 1225189 比例 The ratio of the sum of the 値 to the sum of the parameters 値 of the automatic regression section It is fixed and simple. The restrictions are different, which is a bit complicated. The above operation is an operation example of the static parameter identification mechanism 4. The dynamic parameter identification mechanism 6 can be easily performed, and when the static parameter identification mechanism 4 operates in the following manner, It is possible to obtain a robust prediction that does not reduce the accuracy of the prediction when there is some interference. In the static parameter identification mechanism 4, equation (4) can be expressed by equation (10) in a time zone where no shift operator is used. Q (t) =-aiQ (t-1) -a2Q (t-2) -anQ (tn) + biR (t-1) + b2R (t-2) bmR (tm) (10) Assuming good weather and R (t) = 0. In general, because Part of the sewage field is a mixed sewage system, and sewage containing domestic wastewater and industrial wastewater will flow into the sewage field even when there is no rainfall. In this case, formula (1 1) can be obtained from formula (10). Q (t) --aiQ (t-1) ^ a2Q (t-2) ^ -...- anQ (tn) (11) Actually, the amount of sewage inflow changes moderately. However, for the sake of simplicity, 'assume The inflow of sewage is a fixed inflow qg. Then, the expression (1 1) can be written as:

Qo = (-ai-a2 -...- an) Q〇 (12) In order to keep formula (1 2) well, it is equivalent to the fact that A (1) is 0, where A (1) is automatic The sum of the parameters 値 of the regression part A (q · 1). This is also equivalent to that the automatic regression unit A (q · 1) has an integrator and the zero point of A () is equal to 1. -18- (15) (15) 1225189 When making a prediction by using a model with an integrator in the automatic regression section, the perturbation effect of the fixed chirp can be invalidated. Since the amount of sewage inflow can be regarded as a fixed disturbance, the impact of the amount of sewage inflow can be completely invalidated by conforming to A (1) = 0, and a solid prediction is obtained. However, if A (1) = 0, then the right side of equation (9) is infinite, and therefore, the constraint condition cannot be satisfied. This problem can be solved when A (1) is a very small 値 represented by the formula (1 3). Α (1) = ε, Ο &lt; ε «1 .... (13) Considering the stability of the process that can be checked separately 'e must be positive. From equation (1 3), it can be seen that a solid prediction can be obtained. At the same time, the static parameter recognition mechanism 4 will be reduced to a very practical condition. Generally fixed. Automatic regression section: A &lt; 1) = l + ai + a2 + ··· 〇 &lt; ε ..... << 14) Moving average section: B (l) = bi + b2 + ... bm = k-ε .... &lt; 15 &gt; It is clear from the equations (l4) and (1 5) that when the outflow coefficient k and the constant ε of the positive 値 are very close to zero, the parameter 値 of the automatic regression unit The sum of the sum 値 and the parameter 値 of the moving average part must satisfy equations (1 4) and (1 5). Therefore, the dynamic parameter recognition mechanism 6 only needs to identify the parameters to meet the conditions of "the sum of the parameters 値 of the automatic regression section is fixed" and "the sum of the parameters 値 of the moving average section is fixed", so that it can obtain a very good match Expected parameter identification. -19- (16) (16) 1225189 The static parameter identification mechanism 4 operates as described above. The number of identified parameters determining means 5 determines the number of parameters η and m to be identified in the automatic regression section and the moving average section. The numbers η and m can be determined automatically or manually. The number of parameters η and m can be determined by the following mechanical method. The data stored in the identification / confirmation data storage unit 31 is divided into identification data and confirmation data. Using the identification data and the number of changes n and m 値, the parameters are identified using generally known parameters, such as the Least Squares method. Then, predictions are made by using the identified parameters and confirmation data. The number of identified parameters η and m is one that minimizes the sum of the squared prediction errors of the confirmed data. In general, this method is called the degree of determination method using a loss function. The optimal number of identified parameters η and m can be determined according to a criterion called (Akaike Information Criterion / Information Criterion). Instead of AIC, it can be called MDL (Minimum Description Length). As these methods determine the numbers η and m, dynamic parameters can be identified simultaneously. This identification is performed only on the determination of the optimal number of identified parameters, so that one of the identified parameters does not need to satisfy the limiting conditions determined by the static parameter identification mechanism 4. So these 値 will not be used as the final parameter 参数. Provided that one of the identified parameters satisfies the restriction condition 'to be imposed on the parameter determined by the static parameter recognition mechanism 4, the recognition result can be directly used as the dynamic parameter recognition mechanism 6. The number of parameters η and m can be determined by the following manual method. Retrieve rainfall data and inflow data from the identification / confirmation data storage unit 31. -20- (17) (17) 1225189 and the rainfall data and inflow data will be plotted in time series on the screen of an appropriate display device. The rate of change of inflows in response to changes in rainfall is then estimated. If the inflow quickly responds to the change in rainfall ’, the effect of the automatic regression unit will increase with the increase in the number η, and if η is large, the change in the backward rainfall is predicted, so η = 1 is preferred. Although a solid prediction can be made in the presence of fixed interference as described above, when '1', when there is variable interference, the prediction is not very robust. On the other hand, if the causal relationship between rainfall and inflow is only a qualitative relationship, where the change in inflow does not respond quickly to the change in rainfall, and the inflow increases with rainfall, the use of large The number as the number η is equal to the importance of placing the previous rainfall as a predictor more than the importance of placing the rainfall. Therefore, it is preferable that the number η is increased to a certain degree. When the number? Is large, when there is variable interference, a fairly robust prediction can be made. There is a qualitative relationship between these factors. Table 1 Follow the prediction of η &lt; Dependence between energy and ruggedness Rainfall following performance Anti-interference ruggedness η Fast response At that time, it was strong enough to resist fixed interference Large η Slow response was strong enough to resist fixed and variable interference The relationship shown in Table 1 serves as a criterion for determining the number η. For example, 'determines the number? In the following manner. From the graph obtained by plotting -21-(18) (18) 1225189 rainfall data and inflow data according to time series, evaluate the duration of the impact of rainfall on the inflow. For example, from the graph showing the individual changes in rainfall and inflow over time, you can see how much time it takes for the inflow to return to the inflow level when the radon is good after the end of the rainfall. Based on this assessment of the duration of repeated rainfall events' impact on inflows, the approximate duration of the impact of rainfall on inflows can be determined. For example, suppose that the evaluation is performed in 5 minutes and the holding time of the impact of rainfall inflow is 30 minutes. Then m = 3 0/6 = 5. However, if η &gt; 1, m 値 of 小于 smaller than the graph estimated by displaying rainfall data and inflow data in time series can be used as the final 値 of the number η, because when n &gt; 〇, The regression department continuously corrects forecasts. 'The operation described above is an example of the operation of the dynamic parameter identification mechanism 5. The dynamic parameter identification mechanism 6 specifically determines the number of identified parameters. The number of parameters determined by the determination mechanism 5 that meets the limit conditions determined by the static parameter identification mechanism 4 is determined. This can be achieved, for example, by performing an optimization method called genetic deduction in the following optimization method. The following expression is defined as the evaluation function (the inverse of the adaptability). J = 2 (Q (t) -Q (t)) = (yy) Tfr-y) (16) y: = [Q (i) Q (2) /, &lt; ^ N) r (17) 歹: = [⑽ 你), ··· ， _)] &quot; (18) Among them, the text marked with Λ above is prediction 値, N represents the number of data. Then, it is decided to satisfy a plurality of initial parameters 范围 in the range of 22- (19) 1225189 cases determined by the static parameter recognition mechanism 4. Multiple 値 must be determined so that the total of ai and the total of bi are fixed. This can be achieved by removing one of the parameters of each of ai and bi. When 8! And b! Are defined by using random numbers and the following expressions, ai (i = 2, 3, ... and η) and bi (i = 2, 3, ..., and m) other than 8! And bi are generated. , The total needs to be fixed bj = ε · !! &lt;:-31 = 卜 1-》 丨, i = 2 Then, the 値 generated by using a random number is represented by a binary notation of appropriate size, and Binary puppets serve as the primitive individuals of genetic deduction. By repeatedly accepting the intersection, mutation, and selection of the binary 値, it is determined 値 that minimizes the evaluation function represented by equation (16). In this way, a dynamic parameter identification mechanism is executed. J = (yy) T (yy) = (y-Φθ) τ (y — Φθ) Φη Φΐ2 * VV 1 Φη Φΐ2 · θ / Φ21 〇22 · θ2.] Y \ Φ21 Φ22 θ2 &lt; Qi: = [-ai ,-彐 2,…, -an] T (20) ㈣ q2: = [bi, b2, ..., bn] T (21) where Φ is stored in the identification / confirmation data storage mechanism by proper configuration The time series rainfall data R (t) and time series inflow data Q (t) in 31. The limiting conditions determined by the static parameter identification mechanism 4 are expressed by equations (22), (23), and (24). -23- (20) (20) 1225189 eTA = 1-ε (22) eI02 * (23) e,.: = [U. &quot; 1,] T (24) v — 2 The dynamic parameter identification mechanism 6 can Restrictions on the linear equation of the parameters are formulated into quadratic programming problems. By using a Lagrange method with an indefinite coefficient, for example, the restricted evaluation function is reduced to the following new unrestricted optimization problem, and the restricted optimization method can be solved. J »(y- &lt; D0) T (y-〇e) + \ + ε) + λ2. $: Θ2-ε · k) (25) where λ1 &amp; Λ2 is a Lagrangian indeterminate coefficient. By formulating (25) for 0 part and making? ? ? ? ? . By solving the equation (26), the dynamic parameter identification mechanism 6 can be executed, and the constraint conditions determined by the static parameter identification mechanism 4 can be satisfied. For example, the solution method is described in detail in Imano Hiroshi, Yamashita Hiroshi's "Nonlinear Programming", Nikka Giren (1994). After all the dynamic parameters 决定 are determined by the dynamic parameter recognition mechanism 6, these 値 will be stored in the parameter 値In the storage unit 22. After the prediction system is constructed with the prediction system, the following operations are performed for prediction. The rainfall measuring device and the inflow measuring device 12 will measure the rainfall and The inflow to obtain the data. The data obtained by the rainfall measurement device 11 and the inflow measurement device 12 will be stored in the input and output data storage mechanism 3 in the forecast storage mechanism -24- (21) (21 ) 1225189 3 2. The prediction data storage mechanism 32 will store at least m pieces of rainfall data and n pieces of inflow data. Then the 'dream number 値 storage unit 2 2 will send the parameter 値 to the prediction model of the prediction die cracking device 2 Forming unit 21, and then the rainfall data and inflow data stored in the forecasting data storage mechanism 32 will be transmitted to the forecasting model forming unit 21. Formation unit 2! The prediction model represented by the formation formula (27): Q (&gt; = ^ aiQ (t-ι) -a2Q (t-2) -anQ (tn) + biR (t-1) -fb2R (t- 2) + ... bmR (tm) (27) The rainfall data and the inflow data are given to the prediction model formed by the prediction model forming unit 21, and then the inflow is predicted by using the formula (2 7). By repeating the calculation using the formula (2 7), the long-term inflow can be predicted. When the final prediction is completed, the unnecessary data in the rainfall data and inflow data stored in the forecast data storage mechanism 32 will be Sent to the identification / confirmation data storage mechanism 3. 1. Perform a series of operations in a predetermined prediction period. The prediction system in the first embodiment is different from the construction method of the prediction model based on the conventional black box approximation. The restriction conditions of the parameters determined by the mechanism 4 are used to identify the parameters identified by the dynamic parameter identification mechanism 6. Thus, the prediction system of the present invention has the following effects. (1) The restriction conditions of the parameters determined by the static parameter identification mechanism 4 generally correspond to To the law of physical conservation or similar, and Necessary conditions that must be met by the prediction model. In this way, a prediction model that can satisfy these conservation laws can be constructed with certainty. -25- (22) (22) 1225189 (2) If the data contains abnormal data or is known Prediction model formation method When the data is insufficient when the prediction model is constructed and the prediction model cannot function at all, the parameters will not diverge. The prediction model formation method of the present invention avoids this phenomenon and can form a solid prediction model. (3) No matter what the interference is, a solid prediction can be obtained and a prediction that is not affected by the interference can be obtained. Second Embodiment A second embodiment of the present invention will be described with reference to Figs. 2 to 5, in which portions similar to the first embodiment will be denoted by the same reference numerals and descriptions thereof will be omitted. Fig. 2 shows a prediction system in a second embodiment, which is applied to river water level prediction. The river system 50, that is, the 'operation target of the prediction system shown in FIG. 2' includes a tidal water level measurement device 5 1. An upper pump discharge measurement device 5 2. A drainage pump discharge measurement device 5 3. A drainage pump station River water level measuring device 5 4 and river treatment 5 5. The river system 50 includes regional river treatment 5 5. Causal variables that affect river water levels and river water levels (ie, outcome variables). Figure 3 shows the river system 50 in detail. Referring to FIG. 3, the river system 50 includes a tidal water level measuring device 51, which continuously or at a predetermined time period measures the tidal water level of the river water level in the drainage pumping station 54a that affects the river system 50. Tidal water level measurement device 51 1 The water bay installed in it is connected to the river with -26- (23) (23) 1225189 drainage pump station 54a installed beside it. The upper pump discharge measuring device 52 included in the river system 50 measures the pump discharge of the pump above the river level in the drainage pumping station 54a of the river system 50 continuously or at a predetermined time period. Pump emissions are affected by rainfall in this area. The river system 50 also contains a rainfall measuring device 58. The drainage pump discharge measuring device 5 3 of the river system 50 continuously measures or measures the pump discharge of the drainage pump station 5 4 a which directly affects the river water level in the drainage pump station 54a of the river system 50 at a predetermined period. The prediction system shown in Figs. 2 and 3 is similar in configuration to the prediction system shown in Fig. 1, but the former is applied to the condition prediction of the river system, and the latter is applied to the inflow prediction of the sewage field. The prediction system in the second embodiment is different from the prediction system in the first embodiment in the following two points. One of the two points is that the prediction system includes a first dynamic parameter identification mechanism 61 and a second dynamic parameter identification mechanism 62. Rather than a single dynamic parameter identification mechanism. Another point is that the display device 7 and the marking device 8 are interactively connected to the second dynamic parameter identification mechanism 62. The operation of the prediction system in the second embodiment will be explained. The tidal water level data, the upper pump discharge data, which are measured by the water level measurement device 5 1, the upper pump discharge measurement device 5 2, the drainage pump discharge measurement device 53 and the drainage station water level measurement device 54 in a predetermined period, Drain pump discharge data and drainage station water level data will be given to the input and output data storage mechanism 3 and stored therein in time series. Since no predictive model is constructed at the initial stage, these data will be stored in the identification / confirmation data storage machine -27- (24) 1225189 Structure 31. The first dynamic parameter identification mechanism 61 recognizes the dynamic parameters of the prediction model using a conventional system. The numbers are identified under the assumption that the system has multiple inputs, tidal levels, discharge from upper pump, and discharge from drain pump. However, correct identification was obtained in many cases where the amount of data was insufficient. In this case, parameter identification is performed on each part with Japanese Patent Application / 2002, and River Stage Prediction System. Since when the first dynamic parameter identification mechanism 61 uses the method according to the application number 24231/2002 ^ River Stage Prediction, it is based on, for example, the number of AIC or MDL criteria. Therefore, the number of identified parameters determines the dynamic parameter of the mechanism 5. The identification mechanism 61 operates simultaneously. Since the static parameters need to be determined when the first moving mechanism 61 is executed, the static mechanism 4 and the first dynamic parameter recognition mechanism 61 operate simultaneously when the number of data is small and the conditions for identifying the problem are poor or contain materials. Then, by executing the first dynamic parameter identification mechanism 61, the parameters may be very unreliable. In this case, the static parameter identification mechanism 4 explained again is performed. For example, since the position where the tide level is measured is below the position where the tide level is measured, 'the adjustment between the tide level C (t) and the water level H (t) stored in the identification / confirmation data 31 1 値L. It is assumed that the tide level and identification method in the prediction model forming unit 21, that is, the method described in the dynamic parameter 'Cannot No. 24231' is referred to as the present patent System &quot; to determine the identification of the participant and the identification of the first state parameter. However, the abnormal assets are determined to be static and will be stored in the lower level. All institutions will have river water levels. The relationship between -28- (25) 1225189 is expressed by the following discrete-time transfer function model. D (q 1) = l-fdlq &quot; 1 + d2q &quot; 2+ ... + dnq'n (Automatic regression section) (29) N1 (q'1) = nilqel + na2q ~ 2 +.. · + Nlxnq- m (moving average) (3〇) Water and water When the average tidal water level C0 is continuously input, it is assumed that the average river level is Η 〇

Η ⑽). nii + ni2nn! m c 0 D (l). 1 l + A + dw + dn So, 1 = Co · N (1) / D (1)-Co must be maintained. Therefore, the total ratio of the parameters 自动 of the automatic regression part and the moving average part is expressed by the formula (3 2) Means. Nl (l) _UC0 is good. In combination C0 (32) Basically, for 以外 other than C (t), the formula (3 2) must be good. However, it is practically impossible, so these are shown on average. If the ratio between the sum of the numbers determined by the identification of the first dynamic parameter recognition mechanism 61 is different from the ratio shown in equation (32), then the equation (32) is stored as the static parameter recognition mechanism 4, and The first state parameter recognition mechanism 62 will recognize the parameters again to meet the limiting conditions determined by the static reference mechanism. Similarly, the relationship between the discharge amount P (t) of the drainage pump and the river water level H (t maintains the reference value of the second movement when the table is large) is -29-(26) 1225189, and the discharge amount of the upstream pump P u (t ). The relationship with the river water level Η (t) is determined in advance by the static parameter identification mechanism 4 when necessary. When the drainage pump discharge amount P (t) and the upper pump discharge amount pu (t) remain fixed, the river water level will not be fixed. Assuming that the upper pump discharge Pu (t) is fixed and the drainage pump discharge P (t) = 0, the river water level will continue to rise unless the river floods. If P u (t) = 0 and P (t) are fixed, the river water level continues to decrease until the river dries up. Therefore, the relationship between pump discharge and river water level is considered to have an integral feature. In this case, the following static parameter identification mechanism 4 is possible. For example, if the pump discharge Pu (t) is fixed, the river water level will rise at a fixed rate. The ratio (gradient) s (t) of the rise in river water level to the unit discharge can be expressed by the following expression. Seeing (t) = Jiejia Huica 1) = (1- to _ '(0 ㈣ (34) N2 (qJ) = + where N2 (q ^) is used for the pump discharge Pu (t) The polynomials formed by the coefficient parameters. If the relationship is integrated like this river system, then ^ (ζΓ1) = (1-分 丨) ((分)) must be maintained well. This relationship is substituted into equation (3 3) To obtain formula (3 5). -30- (27) 1225189 s (〇Pu (〇 «

(l-q-1) N2 (q ^) (^ q-) D (q ^ T p, (0

PuO) N2 (q1 Assuming pu (t) = pu 〇 (constant), the rate of rise of this retention is uniform and converges at a fixed 値. Surface; 0 (1) D (l) ^ is obtained in the following manner. · Π (π Μ!-π ㈣ ϋ (1) = ^ q: i) D (q ^ q 1 "= -1 ((^ + 2 (12 +… + ndn) _ N2 (l) n21 ten n22 ten · • • Ten n2p S ~ D (l) * -l- (di + 2d2 + * -4-ndn) The formula (3 9) does not impose restrictions on static parameter identification parameters. Because s represents a single water level The gradient of the curve of the change, so the true time step s for the step when the pump discharge is PuG. The number recognition mechanism 4 with integrated characteristics can be identified in the above method. If the first dynamic parameter is used to identify (35): Guihui and river water (36) (37) (38) (39) The river calculation of the discharge amount determined by the structure 4 must be divided by PuG to implement the static participation agency 6 I decision -31 -(28) (28) 1225189 When the right side of the formula (3 9) and s calculated are very different from each other, the second dynamic parameter recognition mechanism 6 2 will recognize the parameters again to satisfy the formula (39) ° The recognition operation of the second dynamic parameter recognition mechanism 62 will be explained. . As shown in FIG. 2 'the second dynamic parameter identification mechanism 62 will be connected to the display device 7 and the marking device 8. The display device 7 will also be connected to the first dynamic parameter identification mechanism 61. Figures 4 and 5 are examples of display displays Device 7 and marking device 8. Referring to FIG. 4, when the first dynamic parameter identification mechanism 6 is used to identify the dynamic parameter, the predicted river level can be displayed by the display device 7. At the same time, it is stored in the identification / confirmation data storage device. The upper pump discharge data in 3 1 'Drain pump discharge data, tidal water level data and river water level data can be displayed by the display device 7. It can be easily seen from FIG. 4 that these data displayed by the display device 7 have high levels. The prediction accuracy part and the low prediction accuracy part. The marking device 8 separates the low prediction accuracy from the prediction accuracy that needs to be improved as shown in Fig. 4. At the same time, the marking device 8 will infer that it is regarded as The biggest cause of low forecast accuracy (river water level, tidal water level, upper pump discharge or drainage pump discharge) and shows the cause variable. For example, in Figure 4, When the discharge from the upper pump and the discharge from the drain pump are substantially zero, the prediction accuracy is satisfactory; that is, when there is no rainfall, the prediction accuracy is satisfactory. Therefore, it can be known that the rainfall is Reasons for reducing the accuracy of the forecast. Because the tidal water level does not depend on the rainfall at all, the tidal water level will not be the reason for the decrease in the prediction accuracy. -32- (29) (29) 1225189 The prediction accuracy in the state is quite high. 'Uploading pump emissions are hardly the cause of the decrease in prediction accuracy. In a state where the amount of drained chestnuts suddenly starts to increase, the accuracy of the prediction is satisfactory. As a result, it was inferred that the reason for the decrease in prediction accuracy was the coefficient of the river water level (automatic regression coefficient) and the secondary reason was the drainage pump discharge. In this case, the 'river water level is considered to be the main cause and the marking device 8 indicates the river water level as the cause. Therefore, the labeling device 8 labels the portion where the prediction error is desired and the cause of the prediction error. The parameter of the river level specified as the cause is then stored and backed up. Then, the second dynamic parameter recognition mechanism 62 performs recognition to reduce the prediction error. Specifically, the second dynamic parameter identification mechanism 62 optimizes the parameter 値 using a search optimization method such as a genetic deduction method or a restricted quadratic programming method. The second dynamic parameter identification mechanism 6 2 differs from the first dynamic parameter identification mechanism 61 in that the evaluation function is defined not to be used for all prediction errors but only for prediction in the portion marked by the device 8. Errors, and dynamic parameters are identified so as to reduce prediction errors in the marked part. Therefore, in some cases, when the second dynamic parameter recognition mechanism 8 recognizes the parameters again, the prediction error in a portion other than the portion marked by the marking device 8 may increase. The homotopy approximation method can be used to avoid this phenomenon. It uses the parameter 辨识 identified by the first dynamic parameter recognition mechanism 61 as the initial 値 and gradually and continuously changes the initial 値. If the gradually changed predicted parameters are simultaneously displayed by the display device 7, when the approximate predicted parameters are displayed by the display device 7, the operator performing parameter identification can stop the second dynamic parameter identification-33- (30) ( 30) 1225189 Parameter identification operation of mechanism 62. If the method is executed by the second dynamic parameter identification mechanism 62, and the parameter cannot be satisfactorily identified, then the first dynamic parameter identification mechanism 62 will not recognize the parameters of the river water level identified by its original value and another reason variable. It is marked by the marking device 8. For example, drainage pump discharges are labeled as cause variables. Then, the second dynamic parameter identification mechanism 62 executes the same method to identify the parameters. In this way, the prediction error in the portion marked by the marking device 8 will be reduced as shown in Fig. 5, and the prediction accuracy of other portions can be improved. In this way, the prediction model in the prediction mode forming unit 21 is configured. Next, in accordance with the method described in the first embodiment, a prediction is obtained. As is clear from the above description, the prediction system provided with the first dynamic parameter identification mechanism 61 and the second parameter identification mechanism 62 can improve the prediction accuracy of a particular part and can provide a prediction model with strict prediction. Third Embodiment A prediction system according to a third embodiment of the present invention will be described with reference to Figs. 6 and 7. The prediction system in the third embodiment is substantially the same in configuration as the prediction system in the second embodiment shown in FIGS. 2 to 5, but differs from the prediction system in the second embodiment only in that the The parameter selection device 9 is adjusted instead of the marking device 8. In the prediction system shown in Figs. 6 and 7, the second dynamic parameter recognition mechanism 62 performs manual dynamic parameter recognition instead of automatic dynamic parameter recognition. -34- (31) (31) 1225189 The coefficient parameter of the cause variable as shown in Fig. 7 is displayed in the part of the display device 7. Select the cause variable that is estimated to reduce the accuracy of the prediction in a particular place. Select as many pairs of parameters as the cause variable to be adjusted. For example, the coefficient parameters n31 and n34 of the drainage pump discharge volume shown in FIG. 7 are selected for adjustment. Select the two parameters according to the following criteria. In the table shown in Fig. 7, with respect to time prediction, the farther right parameter represents the farther past influence, and the farther left parameter represents the closer to the past. In the table shown in Figure 7, the shorter the distance between the two selected parameters, the greater the impact at a particular time, and the longer the distance, the longer the duration of the impact. The adjustable parameter selection device 9 refers to the data displayed on the display device 7 and selects two parameters. Since the adjustable parameters n31 and n34 are selected, these parameters will be adjusted in pairs. The static parameter identification mechanism 4 sets a restriction condition that the total of the parameter 値 is required to be consistent with the constant. Therefore, when determining 之一 for one of the two parameters, 値 for the other parameter is automatically determined. For example, if the 値 of the parameter n31 changes from -1.2 to-〇 · 8, the 値 of the parameter n34 must change from 0.6 to 〇 · 2. In this way, the parameter 値 can be adjusted to meet the limiting conditions determined by the static parameter identification mechanism 4. When the parameter adjustment operation is performed, the data displayed on the display device 7 is observed visually by predicting the change in the prediction caused by the parameter adjustment. Taking the standards (a) and (b) into consideration, gradually adjusting the parameters, and the display device 7 will display -35- (32) 1225189 as shown in Figure 5 for the final prediction 値. As can be seen from the above description, in the third embodiment, by connecting the adjustable parameter selection device 9 to the second dynamic parameter 6 to adjust the two parameters in pairs, the actual parameter adjustment can be manually improved. Forecast accuracy. Therefore, when the accuracy of local condition prediction is needed, it can be adjusted by manual system. Although it is believed that the parameters are adjusted by the approximation method because the parameters have no physical meaning, even using the black box approximation method can still fine-tune the parameters. The present invention has the following effects. The determination of static constraints, that is, the steady-state control conditions, will be able to identify dynamic models with static model identification and other combinations under constraints. Therefore, using the black box near model can include physical constraints such as the law of conservation of mass, etc. Under the conditions where the available data contains abnormal data, or if sufficient conditions are not available, a reliable prediction model can still be established. It can obtain the parameters of the prediction model using the black box approximation to improve the prediction accuracy. [Brief description of the drawings] FIG. 1 is a prediction system according to the first embodiment of the present invention, FIG. 2 is a prediction system according to the second embodiment of the present invention, and one of the number recognition mechanism pairs in the prediction system is to improve specific acquisition parameters. When it is difficult to take the black box into consideration, the parameters of the present invention are limited to the prediction of the dynamic model recognition method. Even if the data is not fine-tuned manually, the block diagram of the traditional block diagram is -36- (33) (33) 1225189 Figure 3 is a view of the river system; Figure 4 is a display device and a marking device before adjustment; Fig. 5 is a diagram showing an adjusted display device and a marking device; Fig. 6 is a block diagram of a prediction system according to a third embodiment of the present invention; and Fig. 7 is a diagram for explaining the operation of an adjustable parameter selection device. Comparison table of main components 1 Sewage field inflow system 2 Prediction model device 3 Input and output data storage mechanism 4 Static parameter = SIi m Other institution 5 Pass = ίφ. M Other parameter number S Deciding mechanism 6 Dynamic parameter Ξώιι m Other institution 7 Display Device 8 Marking device 9 Mita M Parameter selection device 11 Rainfall measurement device 12 Inflow measurement device 13 Rainfall-inflow processing 2 1 Prediction model formation unit 22 Prediction model formation unit 3 1 ΞώΞι m Don't / Confirm data storage Institution 32 Prediction data storage mechanism-37- (34) 1225189 50 River system 5 1 Tidal water level measurement device 52 Upper pump discharge measurement device 52a Upper pump station 53 Drain pump station discharge measurement device 54 Drain pump discharge Measuring device 54a Drainage pump station 55 River treatment 58 Rainfall measuring device 6 1 First development Number = ili m respectively mechanism 62 second dynamic parameter identification mechanism -38- Ξώΐΐ

Claims (1)

(1) (1) 1225189 Pick up and apply for patent scope 1 · A prediction system for predicting the prediction target, including: a prediction model forming unit for predicting the prediction target, including a prediction model with multiple parameters; parameter 値 storage A unit for storing multiple parameters; an input and output data storage mechanism for storing time series input and output data for prediction; a static parameter identification mechanism, which determines the parameters to be added to the parameters based on the fixed input and output data Restriction conditions; and dynamic parameter identification mechanism, based on the non-fixed input and output data stored in the input and output data storage mechanism to determine the parameters 値 that meet the limits determined by the static parameter identification mechanism, and transmit the parameters 値 to Parameter 値 storage mechanism. 2 · The prediction system according to item 1 of the patent application scope, wherein the prediction model forming unit includes an automatic regression section and a moving average section, and the automatic regression section linearly uses the time series output data as a prediction model, and the moving average section performs time series The input data acts linearly; and the static parameter identification mechanism imposes restrictions on the parameter 値, so that the ratio bb / aa is fixed, where aa is the sum of the coefficients 自动 of the automatic regression section, and bb is the moving average section. The sum of the coefficients 値. 3. The prediction system of item 2 of the scope of patent application, wherein the static parameter identification mechanism will impose restrictions such that the total aa of the parameter 値 of the automatic regression unit is a fixed positive number very close to zero and aa = e means £ &lt; &lt; 1. -39- (2) (2) 1225189 4 · For example, the prediction system for the scope of patent application No. 2, in which the dynamic parameter identification mechanism, the static parameter identification mechanism, according to the requirements, use the optional optimization mechanism and stored in the input and The non-fixed input and output data in the output data storage mechanism must be fixed in the previously determined parameters of the automatic regression section and the moving average section. The parameter 値 must be fixed. 5 · A prediction system for predicting the prediction target, including: a prediction model forming unit for predicting the prediction target, including a prediction model with multiple parameters; a parameter 値 storage unit for storing multiple parameters 値; input And output data storage mechanism for storing time series input and output data collected for prediction in a predetermined period of time; the table-static parameter identification mechanism 'according to the non-fixed input and output data stored in the input and output data storage mechanism, Determine the parameters 値; the static parameter identification mechanism, based on the real or fictional fixed input and output data, determines the constraints imposed on the parameters; and the second dynamic parameter identification mechanism, based on the non-fixed stored in the input and output data storage mechanism Input and output data, adjust the parameter 値 determined by the first dynamic parameter identification mechanism so that the parameter 値 is within a range that meets the limit conditions determined by the static parameter identification mechanism, and transmit the parameter 给 to the parameter 値 storage mechanism. 6 · If the prediction system of the scope of patent application No. 5 further includes: a display device for simultaneously displaying the input and output data stored in the input and output data storage mechanism and corresponding to the input and output data and from -40- ( 3) (3) 1225189 Prediction data provided by the prediction model forming unit; and a marking device for marking the part of the prediction model that needs to reduce the prediction error, and the variable considered to be the cause of the error; of which, the second dynamic parameter The identification mechanism will adjust the parameters of the cause variable marked by the marking device to reduce the prediction error in the portion marked by the marking device. 7 · If the prediction system of the scope of patent application No. 5 further includes: a display device for simultaneously displaying the input and output data stored in the input and output data storage mechanism and a unit corresponding to the input and output data and formed by a prediction model Provided prediction data; and adjustable parameter marking and display devices for indicating the part of the prediction model whose prediction error needs to be reduced relative to the data displayed by the display device, and the variables that are considered to cause the errors, and, Used to select at least two parameters in the cause variable for manual adjustment. -41-