Inference rule determining method and inference device
Download PDFInfo
 Publication number
 USRE36823E USRE36823E US08542852 US54285295A USRE36823E US RE36823 E USRE36823 E US RE36823E US 08542852 US08542852 US 08542852 US 54285295 A US54285295 A US 54285295A US RE36823 E USRE36823 E US RE36823E
 Authority
 US
 Grant status
 Grant
 Patent type
 Prior art keywords
 inference
 input
 output
 network
 rule
 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.)
 Expired  Lifetime
Links
Images
Classifications

 G—PHYSICS
 G06—COMPUTING; CALCULATING; COUNTING
 G06N—COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
 G06N7/00—Computer systems based on specific mathematical models
 G06N7/02—Computer systems based on specific mathematical models using fuzzy logic
 G06N7/04—Physical realisation
 G06N7/046—Implementation by means of a neural network

 G—PHYSICS
 G05—CONTROLLING; REGULATING
 G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
 G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
 G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
 G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
 G05B13/0285—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks and fuzzy logic

 G—PHYSICS
 G06—COMPUTING; CALCULATING; COUNTING
 G06N—COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
 G06N5/00—Computer systems utilising knowledge based models
 G06N5/04—Inference methods or devices
 G06N5/048—Fuzzy inferencing
Abstract
Description
This .Iadd.is a reissue of application Ser. No. 07/904,690, filed Jun. 26, 1992, now U.S. Pat. No. 5,255,344, which .Iaddend.is a continuation of application Ser. No. 07/459,815, filed as PCT/JP89/00504, May 19, 1989, now U.S. Pat No. 5,168,549.
The present invention relates to inference rule determining methods and inference devices in controlling apparatus, in the inference of an expert system, and in pattern recognition based on input data.
In order to describe a conventional technique, first, the basic fuzzy inference will be outlined by taking as an example fuzzy control used in apparatus control, etc.
In a control system which relates to the evaluation of human beings, the operator can determine a final manipulated control variable using a variable which the operator has determined subjectively and/or sensuously, for example, "large", "middle", "tremendously", or "a little" (which is hereinafter referred to as a fuzzy variable). In this case, the operator determines the manipulated variable from the input variables on the basis of his control experience. An inference device using fuzzy control assigns an input fuzzy number to an IF part of an inference rule in accordance with the inference rule of "IF . . . THEN . . . " type and determines an output fuzzy number of the THEN part from a fitting grade (membership value) indicative of the extent to which the inference rule is satisfied. The actual manipulated variable can be obtained by taking the center of gravity value, etc., of the output fuzzy number.
One of the conventional control methods using fuzzy inference is fuzzy modeling disclosed, for example, in GeanTaek Kang, Michio Sugano; "fuzzy modeling" SOCIETY OF INSTRUMENT AND CONTROL ENGINEERS PAPERS, Vol. 23, No. 6, pp. 650652, 1987. In the control rule of the fuzzy modeling, the IF part is constituted by a fuzzy proposition and the THEN part is constituted by a regular linear equation between inputs and outputs. If a timing lag of first order tank model is considered, for example, the control rule among a control error e, its change in error de and a control output (manipulated variable) u is given by
If e is Zero and de is Positive Medium
Then u=0.25 e+1.5 de
A plurality of such inference rules are prepared, all of which are referred to as control rules Zero, Positive Medium, etc., are each a label or a fuzzy variable (fuzzy number) used to described the rules. FIG. 1 illustrates one example of a fuzzy variable. In FIG. 1, NB denotes Negative Big; NM, a negative Medium; NS, a Negative Small; ZO, a Zero; PS, Positive Small; PM, a Positive Medium; and PB, Positive Big, A function indicative of a fuzzy number F on X is referred to as a membership function μ_{f} () and the function value of x^{0} is referred to as a membership value μ_{F} (X^{0}). The general form of the control rules is given by ##EQU1## where R^{5} indicates a s^{th} rule; x_{j}, an input variable; A_{j} ^{s}, a fuzzy variable; y^{s}, an output from the s^{th} rule; and c^{s}, a THEN part parameter. The result of inference for an input (x_{1} ^{0}, x_{2} ^{0}, . . . x_{m} ^{0}) is given by ##EQU2## where n is the number of rules, and w^{s} is a fitting grade at which the input (x_{1} ^{0}, x_{2} ^{0}, . . . , x_{m} ^{0}) is adapted to the IF part of the S^{th} rule W^{s} is given by ##EQU3## where the membership value in the x^{0} in the fuzzy variable A_{j} ^{s} is μ_{Aj} ^{s} (xj). The identification of a fuzzy model includes a twostage structure, namely, identification of the structure of the IF and THEN parts and identification of the IF and THEN parts. The conventional identifying process includes the steps of (1) changing the fuzzy proposition of the IF part to a proper proposition, (2) changing W^{s} in a constant manner, (3) searching only the actually required ones of the input variables of the THEN part using a backward elimination method, (4) calculating parameters of the THEN part using the method of least squares, (5) repeating the steps (2)(4) to determine an optimal parameter, (6) changing the fuzzy proposition of the IF part and (7) returning to the step (2) where an optimum parameter is repeatedly arched under the conditions of a new fuzzy proposition. Namely, this method can be said to be a heuristic methodlike identifying algorithm.
A conventional inference device includes a fuzzy inference device, for example, shown in FIG. 2 in which reference numeral 1a denotes a data input unit (including a measured value and a value evaluated by human being); 2a, a display command unit; 3a, a fuzzy inference operation unit; 4a, an inference result output unit; and 5a, a display. The display 2a is composed of a keyboard, and the fuzzy inference operation unit 3a is composed of a digital computer. The input data to the digital input unit 1a is subjected to inference operation at the fuzzy inference operation unit 3a. Thus, the operation unit 3a outputs the result of the inference and simultaneously displays a list of inference rules, a list of fuzzy variables and the states of use of various inference rules on the display 5a. In a conventional inference device such as that shown in FIG. 2, the inference rules of the fuzzy inference rules and the fuzzy variables as input data are fixed as constant values in the fuzzy inference operation unit 3a and have no function of changing the fuzzy variables.
Since an algorithm of determining a membership function is based on heuristic method in the conventional inference rule determining method as well as in the fuzzy modeling, it is complicated and the number of parameters to be determined is very large. Therefore, optimal inference rules cannot be obtained easily at high speed.
Since an influence device such as that shows in FIG. 2 has no function of learning inference rules, the characteristic of the inference variable changes with time, so that it cannot cope with the situation that the inference accuracy would be deteriorated according to the inference rules set initially. Assume, for example, that there is the inference rule "If it is hot, then control diarie a so as to be in level B." Since the obscure concept "hot" varies from person to person as well as season to season, satisfactory control cannot be achieved unless the inference device has a learning function and can change the inference rules adaptively in accordance with situations under which the device is used. Furthermore, there are actually many cases where nonlinear inference is required, so that a method of performing linear approximation described with reference to the above conventional example has a limit to improvement in the inference accuracy.
In order to solve the above problems, it is a first object of the present invention to provide an inference rule determining process for automatically determining a fuzzy number contained in the inference rules described in the form "IF . . . THEN. . ." using a learning function without relying upon experience.
It is a second object of the present invention to provide an inference device which is capable of coping with an inference problem at high speed even if the problem is nonlinear, using the nonlinearity of a neural network model. In the inference device, a membership value determiner constituting the neural network infers from the input variables a membership value corresponding to the IF part of each inference rule. The individual inference quantity determiner of the inference device infers from the input variables an inference quantity corresponding to the THEN part of the inference rule. The final inference quantity determiner makes a synthetic determination based on the inference quantity and the membership value inferred for each of the rules for the input variables and obtains a final inference quantity. The present invention is the inference device which functions as described above.
FIG. 1 illustrates fuzzy variables;
FIG. 2 is a schematic of a conventional inference device;
FIGS. 3 and 4 are schematics of neural network models which constitute the membership value determiner and the individual inference quantity determiner;
FIG. 5 is a schematic of a linear signal processor;
FIG. 6 is a schematic of a nonlinear signal processor;
FIG. 7 is an inputoutput characteristic diagram of a threshold processor;
FIG. 8 illustrates a process for automatically determining a membership function using a neural network model;
FIG. 9 is a schematic of an inference device of one embodiment in the present invention;
FIG. 10 is a schematic of input/output data used for describing the operation of the FIG. of embodiment 1;
FIG. 11 illustrates the membership value of each learning data in the IF part of the FIG. 9 embodiment;
FIG. 12 illustrates an inference error produced when the number of input variables is reduced;
FIG. 13 illustrates the result of inferring the inference quantity of the FIG. 9 embodiment;
FIG. 14 illustrates the result of the inference according to the present invention applied to the problem of inferring the COD density in Osaka Bay.
The features of the inventive method lies in sequentially determining membership functions using a learning function. Various methods of realizing the learning function are conceivable. In a firstembodiment, a learning algorithm of a neural network model is used. Before description of the embodiment, a neural network model and a neural network model learning algorism will be described.
The neural network model is a mathematical network which has gotten a hint from the combination of cranial neural cells. Inference rules can be determined without relying on experience by determining the intensity of connection of units constituting the network using sequential leaning.
The network of FIG. 3 is a kind of a neural network model. In FIG. 3, reference numeral 100 denotes a multiinputsingle output signal processor; and 101, input terminals of the neural network model. FIG. 4 shows an example of a 3stage structured neural network model which receives four input variables and output a single output value. There is no interconnection in each stage and a signal is propagated to an upper stage only. FIG. 5 specifically illustrates the structure of a linear signal processor which basically includes a linear computation section of the multiinputsingle output signal processors 100 constituting the neural network model. In FIG. 5, reference numeral 1001 denotes each of input terminals of the multiinputsingle output signal processor 100; 1002, a memory which stores weighting coefficients given to the corresponding inputs from the input terminals 1001; 1003, multipliers which multiply the inputs from the input terminals 1001 by the corresponding weight coefficients in the memory 1002; and 1004, an adder which adds the respective outputs from the multipliers 1003. Namely, the multiinputsingle output signal processor 100 of FIG. 5 calculates
y=Σcpx.sub.i
where x_{i} is an input value to the corresponding one of the input terminals 1001 and c_{i} is the corresponding one of the weight coefficients stored in the memory 1002. FIG. 6 specifically illustrates the structure of a nonlinear signal processor, which performs a nonlinear operation, of the multiinputsingle output signal processors 100 which constitute the neural network model. In FIG. 6, reference numeral 1000 denotes the linear signal processor described with reference to FIG. 5; and 2000, a threshold processor which limits the output from the linear signal processor to a value in a predetermined range. FIG. 7 illustrates the inputoutput characteristic of the threshold processor 2000. For example, the input/output characteristic of the threshold processor 2000 which limits the output from the linear signal processor 1000 to within the range (0, 1) is represented mathematically by
OUT=1/(1+exp(IN))
where IN and OUT are the input and output to and from the threshold processor 2000.
The input/output relationship in the neural network is represented by
Y=NN(x) (1)
In the description of the embodiment, the model size is represented by the klayer μ_{0} x μ_{1} x . . . x μ_{k}  where μ_{i} is the number of neuron models of each of the input, middle and output layers. The above refers to a general description of the neural network model.
The learning algorithm of the neural network model will now be described. Various learning, algorithms are proposed, one of which is a backpropagation (D. E. Rumelhart, G. E. Hinton and R. J. Williams, "Learning Representations by BackPropagating Errors," Nature, Vol. 323, pp. 533536, Oct. 9, 1986). The mathematical verification and formulation of this algorithm are left to references, but it is basically described as follows. First, a multiplicity of sets of input variables as learning data and percentage values belonging to the respective inference rules inferred by other means are prepared. Input data is input to the neural network model. Initially, an appropriate value, for example a random number value, is input in the memory 1002. The error between the output value from the neural network model and the percentage value belonging to the inference rules is calculated. Since the error is a function of the weight coefficients in the memory 1002, the weight coefficient is corrected in the direction of weight differentiation with respect to the error. Namely, the error is minimized and leaning is performed by correcting the weights sequentially in accordance with the Steepest Decent Method. This is the summary of the learning method using Backpropagation Algorithm.
An inference rule determining process and the inference quantity calculating process according to the present invention will be descried using the neural network model and leaning algorithms.
Only input variables having high correlation to an inference value are selected and unnecessary input variables are removed. If the determination is performed, for example, using the neural network model, first, respective input variables x_{j} where j=1, 2, . . . k are assigned to the neural network model and learning is performed such that an inference value y_{i} is obtained. Similar learning is performed by decrementing the number of input variables by one. Similarly, the backward elimination method is repeated. The difference between each output of the neural network model which has completed the corresponding learning and the inference value y_{i} is compared using the error sum of squares as an evaluation index, and only the input variables (X_{j} where j=1, 2, . . . m, where m≦k) to the neural network model present when the error is minimized are selected.
The input/output data (x_{i}, y_{i}) is divided into n_{t} data segments for identifying the structure of in inference rule (hereinafter described as TRD) and n_{o} inference rule evaluation data segments (hereinafter described as CHD) where n=n_{t} +N_{o}.
Each TRD is divided into optimal r subsegments using regular clustering method. Each set of learning data segments to which the TRD is divided is R^{s} where s=1, 2, . . . , r, and the learning data on R^{s} is (x_{i} ^{s}, Y_{i} ^{s}) where i=1, 2, . . . , (n_{t})^{s} where (n_{t})^{s} denotes the number of data of the TRD in each R^{S}. The rdivision of an mdimensional space means that the number of inference rules is r.
The IF part structure is identified. X_{i} is an input value to the input layer of the neural network model and the output value from the output layer is given by ##EQU4## A neural network model is learned such that the grade w_{i} ^{s} of each learning data (x_{i}) belonging to each R^{s} is inferred. The output value from the neural network model after learning is the membership value of the IF part. Namely, the membership function is given by
μ.sub.A.sup.s (x.sub.i)=w.sub.i.sup.s
where i=1, 2, . . . , n
The structure of the THEN part is identified. The THEN part structure model of each inference rule is A represented by the inputoutput relation (1) of the neural network model and the input values x_{i} ^{s}, . . . , x_{im} ^{s} where i=1, 2, . . . , (n_{i} ^{s}) and the output value y_{i} ^{s} are allocated. A neural network model which infers the inference quantity by learning is identified. Substituting the CHD input values x_{il} ^{s}, . . . , x_{im} ^{s} where i=1, 2, . . . , n_{c} into the resulting neural network model and the error sum of squares O^{s} is calculated by ##EQU5## There is another idea to calculate θ_{m} ^{s} with the weight; this is ##EQU6##
Only the input variables contributing to the inference quantity inferring equation of the THEN part of each inference rule are selected using variable reducing method. Only any one of m input variables is removed and a neural network model is then identified for each THEN part using the TRD similarly to step 5. The inferred error sum of squares θ_{m1} ^{sp} of the inference quantity obtained when the CHD is used is calculated as follows: ##EQU7## From the above expression, it will be seen to what extent the removed input variable is important to determine the THEN part. If
θ.sub.m.sup.s >θ.sub.m1.sup.sp,
the input variable x^{p} is discarded because it is considered to be not important.
An operation similar to step 5 is performed by using the m remaining input variables. Subsequently, steps 5 and 6 are repeated and the calculation is stopped when no equation (2) holds for all the input variables. The optimal neural network model is the one where O^{s} minimum.
The steps 17 determine the IF and THEN parts for each inference rule and the identification of the structure of the fuzzy inference rule ends.
The quantity of inference y_{i} is obtained by ##EQU8## where y^{s} shows a presumed value obtained by substituting CHD into the optimal neural network model obtained at step 7.
If the neural network model includes a collection of nonlinear characteristics of FIG. 7, the inference of a nonlinear object is automatically possible by this learning process, of course.
Thus, if the inference rule is determined by the inventive process, the inference rules are obtained easily and accurately without using the heuristic method since the inference rules are determined sequentially using the leaning function even if the object to be inferred is in a nonlinear relationship or in an obscure input/output relationship.
The inventive process described above will now be described using specific experimental values. In the particular experiment, the inference rule and the quantity of inference were determined using simple numeral examples shown in Tadashi Kondo, Revised GMDH Algorithm Estimating Degree of the Complete Polynomial, SOCIETY OF INSTRUMENT AND CONTROL ENGINEERS PAPERS, Vol. 22, No. 9, pp. 928934, 1986. The process for determining the inference rule will be described below.
FIG. 10 shows input/output data where data numbers 120 denote data for determining the structure of an inference rule (TRD) used in learning and data numbers 2140 denote evaluation data (CHD). Therefore, n_{t} =n_{c} =20 and m=2. Input variables were selected by learning 15,000 times in a threelayered (3×3×3×1) neural network model. The results are shown in Table 1.
TABLE 1______________________________________ ERROR SUM OF SQUARES______________________________________when all variables were used 0.0007when x.sup.1 was removed 0.3936when x.sup.2 was removed 0.1482when x.sup.3 was removed 0.0872when x.sup.4 was removed 0.0019______________________________________
The error sum of squares is not so much influenced even if the input variable x^{4} is added or otherwise removed. Therefore, it is determined that it does not virtually influence the determination of the membership function and thus is not used in the subsequent experiments.
The TRD is divided using the regular clustering method. As the result of the clustering, each learning data is dividable into two segments as shown in Table 2.
TABLE 2______________________________________INFERENCE RULES LEARNING DATA NUMBERS______________________________________R.sup.1 1, 2, 3, 4, 5, 11, 12, 13, 14, 15R.sup.2 6, 7, 8, 9, 10, 16, 17, 18, 19, 20______________________________________
In order to infer with the value of W^{s} ε[0,1] a proportion W_{i} ^{s} ={0, 1} belonging to leaning data (x_{i}, y_{i}) where i=1, 2, . . . , 20, the threelayered [3×3×3×2] neural network model was learned 5,000 times to obtain the fuzzy number A^{s} of the IF part. FIG. 11 illustrates the fuzzy number A^{1} of the IF part for the inference rule R^{1} at this time. The input data in Table 2 where μ_{R} '(x_{i}, y_{i})=W_{i} ^{1} >0 was used as the learning data.
Similarly, the fuzzy number of the IF part for R^{2} was calculated.
An inference quantity presuming equation of the THEN part in each inference rule was calculated. The threelayered [3×8×8×1] neural network model was learned 20,000 times and θ_{4} was obtained as shown in Table 3.
TABLE 3______________________________________REFERENCE RULE 1: O.sub.4 27.86REFERENCE RULE 2: O.sub.4 1.93______________________________________
θ^{s} is calculated where any one input variable is removed from the structure model of the THEN part of the influence rule R^{s}. A threelayered [2×8×8×1] neural network model was learned 10,00020,000 times. As a result the error sums of squares of FIG. 12 were obtained for the inference rules R^{1} and R^{2}, respectively. The comparison between steps 5 and 6 for each inference rule shows that ##EQU9##
Therefore, the neural network model at step 5 is used as the THEN part model for inference rule 1. For inference rule 2, calculation is further continued and is ended at the repetitive calculations at the second stage. (The neural network model with (x_{2}, x_{3}) inputs is used as the THEN part. The resulting fuzzy model is given by ##EQU10## FIG. 13 illustrates the results of the calculation of y^{0} in equation (3) using an inference device which executed the rules thus obtained.
While in the particular embodiment the weighted center of gravity defined by equation (3) was employed as the quantity of inference obtained as the results of the fuzzy inference, a process for calculating the central value or maximum value may be used instead. While in the particular embodiment the algebraic product was used in the calculation equation of w^{s}, a fuzzy logical operation such as a miniature calculation may be used instead of the algebraic production.
In the steps of the process for determining the inference rule, the step of determining a nonlinear membership function automatically will now be described in more detail.
It is natural that the same reference rule will be applied to similar input data. At step 3, learning data is clustered (FIG. 8(a)). These classes correspond to inference rules in onetoone relationship. The inference rule is, for example, "IF x_{1} is small and x_{2} is large, THEN . . .". Three such inference rules R_{1} R_{3} are shown in FIG. 8(a). One hundred percent of each of the inference rules is obeyed at typical points (a set of input data) in these classes, but the data arc more influenced by a plurality of such inference rules nearer the boundary. A function indicative of this percentage is the membership function, the form of which is as shown in FIG. 8(c), for example, which is a view of the membership function as viewed from above and the hatched portion is an area where the membership functions intersect.
A step 3, 4 is for preparing such membership functions. A neural network model of FIG. 8(b) is prepared, variable values (x1 and x2 in FIG. 8(b) of the IF part are allocated to the input while 1 is allocated to a rule number to which the input value belongs and 0 is allocated to rule numbers other than the former rule number in the output. Learning is then performed. One of the important features of the neural network model is that a similar input corresponds to a similar output. While in learning the points x of FIG. 8(a) alone are used, the same output is obtained at points near the points x, namely, the output shows that the same inference rule is obeyed. The neural network model which has finished learning outputs optimal balance percentages belonging to the respective inference rules when it receives data in the vicinity of the boundaries of the respective inference rules. These are the membership values. In other words, the neural network model includes all the membership functions of the respective inference rules.
While in the particular embodiment of the inventive process the membership functions of fuzzy inference rules are determined using the learning algorithm of the neural network, other learning algorithms may be used, of course. As will be easily imagined from the fact that the BackPropagation algorithm described in the algorithms of the neural network is based on the classic steepest descent method, many learning algorithms will be easily conceived which are used in the field of nonlinear minimization problems such as Neuton's method and conjugate gradient method. It is commonsense that in the field of pattern recognition a learning method is used in which a standard pattern which is used for recognition is caused to sequentially approach the input pattern. Such sequential learning process may be used.
While in the step 5 of the embodiment the THEN part has been illustrated as being constructed by a neural network, it may be expressed using another process. For example, if fuzzy modeling such as that described in a conventional method is used, the THEN part will differ from that of the first embodiment in terms of nonlinear expression. However, the essence of the present invention will not be impaired which automatically determines the fuzzy number of an inference rule sequentially using the learning function if a learning function such as a neural network model is incorporated in the IF part.
In the particular embodiment of the present invention, the expression "automatic formation of a membership function of a fuzzy inference rule" is used. This special case corresponds to a production rule used in a classic expert system. In such a case, the expression "automatic formation of the reliability of a inference rule" would be more intelligible. However, this is only a special case of the embodiment of the present invention.
A second embodiment of the invention which performs inference on the basis of the inference rule obtained by the inventive method will now be described. FIG. 9 is a schematic of the first embodiment of the inference device according to the present invention. In FIG. 1, reference numeral 1 denotes a membership value determiner which calculates the membership value of the IF part of each inference rule; 2, an individual inference quantity determiner which infers a quantity of inference to be calculated at the THEN part of each inference rule; and 3, a final inference quantity determiner which determines the final quantity of inference from the outputs of the individual inference quantity determiner 2 and the membership value determiner 1. Reference numerals 212r denote the internal structure of the individual inference quantity determiner 21 which determines an inference quantity in accordance with the respective inference rules 1r. The determiners 1 and 212r each have a multistage network structure such as shown in FIG. 3.
If input values x_{1} x_{m} are input to the inference device of the present invention, the membership value determiner 1 outputs the respective percentages in which the corresponding input values satisfy the respective inference rules. The determiner 1 has beforehand finished its learning in accordance with the steps described in the inference rule determining process of the present invention. The individual inference quantity determiner 21 outputs in inference quantity when those input values obey the inference rule 1. Similarly, the individual inference quantity determiners 212r output inference quantities when they obey inference rules 2r. The final inference quantity determiner 3 outputs a final inference result from the outputs of the determiners 1 and 2. The process for the determination uses the equation (3), for example.
As will be clear from FIGS. 6 and 7, the input/output relationship of the multiinputsingle output signal processor 100 which constitutes the neural network model is nonlinear. Therefore, the neural network model itself can perform nonlinear inference. Thus, the inference device according to the present invention realizes a high inference accuracy which cannot be achieved by linear approximation.
High inference accuracy of the present invention will be illustrated using a specific application. The application refers to the inference of the COD (Chemical Oxygen Demand) density from five parameters measured in Osaka Bay from April, 1976March, 1979. The five measured parameters are (1) water temperature (°C.), (2) transparency (m), (3) DO density (ppm), (4) salt density (%), and (5) filtered COD density (ppm). In this data, 32 sets of data obtained from Apr., 1976 to Dec., 1978 were used for learning and the remaining 12 sets of data were used for evaluation after the leaning process. FIG. 14 illustrates the result (solid line) estimated by the present invention and the actual observed values of COD density (broken lines). It will be understood that the presumed result is a very good one.
As described above, the first embodiment of the inventive inference device includes a membership value determiner constituted by a plurality of multiinputsingle output signal processors connected in at least a network such that it has a learning function and is cable of determining a quantity of inference corresponding to the input variables in a nonlinear relationship.
A second embodiment of the inventive inference device is capable of performing approximate inference by storing in memory the beforehand obtained input/output relationship of a predetermined neural network model, and referring to the memory instead of execution of neural network model directly each time this execution is required. While in the first embodiment the membership value and individual inference quantity determiners are described as having a neural network model structure, an inputoutput correspondence table may be prepared beforehand from a neural network model determined by a learning algorithm if the input variables are obvious beforehand and/or if close approximation is permissible. Although the inventive method cannot be incorporated so as to have a learning function in the second embodiment which functions as a substitution for the membership value determiner or the individual inference quantity determiner by referring to the correspondence table during execution of the inference rule, an inexpensive smallsized inference device is realized because no circuit structure such as that shown in FIG. 5 is needed. The inference corresponding to nonlinearity can be performed as in the first embodiment of the inference device of the present invention.
As described above, according to the second embodiment, at least one of the membership value determiner and the individual inference quantity determiner is composed of the memory which beforehand stores the input/output characteristic of the inference unit constituted by a plurality of multiinputsingle output processors connected to at least the network, and the memory referring unit. Thus, the inference device has a simple circuit structure for determining an inference quantity corresponding to input variables in a nonlinear manner.
As described above, according to the present invention, an inference rule is prepared which has a learning function, forms an inference rule automatically and performs the best inference. According to the inventive inference device, inference is performed with high accuracy even if an object to be inferred is nonlinear, so that the device is valuable from a practical standpoint.
Claims (4)
Priority Applications (4)
Application Number  Priority Date  Filing Date  Title 

JP12435488  19880520  
JP63124354  19880520  
US07904690 US5255344A (en)  19880520  19920626  Inference rule determining method and inference device 
US08542852 USRE36823E (en)  19880520  19951013  Inference rule determining method and inference device 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

US08542852 USRE36823E (en)  19880520  19951013  Inference rule determining method and inference device 
Related Parent Applications (1)
Application Number  Title  Priority Date  Filing Date  

US07904690 Reissue US5255344A (en)  19880520  19920626  Inference rule determining method and inference device 
Publications (1)
Publication Number  Publication Date 

USRE36823E true USRE36823E (en)  20000815 
Family
ID=27314911
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

US08542852 Expired  Lifetime USRE36823E (en)  19880520  19951013  Inference rule determining method and inference device 
Country Status (1)
Country  Link 

US (1)  USRE36823E (en) 
Cited By (4)
Publication number  Priority date  Publication date  Assignee  Title 

US6338051B1 (en) *  19980430  20020108  Samsung Electronics Co., Ltd.  User preference modeling method using fuzzy networks 
US6654763B2 (en)  20010614  20031125  International Business Machines Corporation  Selecting a function for use in detecting an exception in multidimensional data 
US20070260357A1 (en) *  20040609  20071108  Jorg Issberner  Method for the Production of Hydrophilic Polymers and Finishing Products Containing the Same Using a ComputerGenerated Model 
US20150331921A1 (en) *  20130123  20151119  Hitachi, Ltd.  Simulation system and simulation method 
Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

JPS61234405A (en) *  19850410  19861018  Hitachi Ltd  Digital control system 
US4730259A (en) *  19850301  19880308  Gallant Stephen I  Matrix controlled expert system producible from examples 
US4773099A (en) *  19851010  19880920  The Palantir Corporation  Pattern classification means for use in a pattern recognition system 
US4837725A (en) *  19851022  19890606  Omron Tateisi Electronics Co.  Fuzzy membership function circuit 
JPH0293940A (en) *  19880930  19900404  Omron Tateisi Electron Co  Fuzzy computer 
Patent Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

US4730259A (en) *  19850301  19880308  Gallant Stephen I  Matrix controlled expert system producible from examples 
JPS61234405A (en) *  19850410  19861018  Hitachi Ltd  Digital control system 
US4773099A (en) *  19851010  19880920  The Palantir Corporation  Pattern classification means for use in a pattern recognition system 
US4837725A (en) *  19851022  19890606  Omron Tateisi Electronics Co.  Fuzzy membership function circuit 
JPH0293940A (en) *  19880930  19900404  Omron Tateisi Electron Co  Fuzzy computer 
NonPatent Citations (22)
Title 

"Fuzzy Modeling" Society of Instrument and Control Engineers Papers, vol. 23, No. 6, pp. 650652, 1987; GeanTaek Kang, Michio Sugeno. 
"Revised GMDH Algorithm Estimating Degree of the Complete Polynomial", pp. 928934, 1986. 
C.G. Looney, "Learning in ThreePlex Fuzzy Logic Networks," Proceedings of the ISMM International Symposium, XP000566865, Feb. 13, 1988, pp. 157160. 
C.G. Looney, Learning in Three Plex Fuzzy Logic Networks, Proceedings of the ISMM International Symposium, XP000566865, Feb. 1 3, 1988, pp. 157 160. * 
D. E. Rumelhart, G.E. Hinton and R. J. Williams, "Learning Representations by Backpropergating Errors", Nature, vol. 323, pp. 523536, Oct. 9, 1986. 
D. E. Rumelhart, G.E. Hinton and R. J. Williams, Learning Representations by Back propergating Errors , Nature, vol. 323, pp. 523 536, Oct. 9, 1986. * 
Fuzzy Modeling Society of Instrument and Control Engineers Papers, vol. 23, No. 6, pp. 650 652, 1987; Gean Taek Kang, Michio Sugeno. * 
Hayashi et al., "NeuralNetworkDriven Fuzzy Reasoning Model", Proc. 28th SICE Annul.Conf., Jul. 1989, 13431346. 
Hayashi et al., Neural Network Driven Fuzzy Reasoning Model , Proc. 28 th SICE Annul.Conf., Jul. 1989, 1343 1346. * 
Isao Hayashi, et al, "Formulation of Fuzzy Reasoning by Neural Network", 4th Fuzzy Symposium (Tokyo, May 3031, 1988), pp. 5560. 
Isao Hayashi, et al, Formulation of Fuzzy Reasoning by Neural Network , 4th Fuzzy Symposium (Tokyo, May 30 31, 1988), pp. 55 60. * 
J. Voevodsky, "Plato/Aristotle: A Neural Net Knowledge Processor," IEEE First International Conference on Neural Networks, XP002058597, Jun. 2124, 1987, San Diego, CA, pp. 399408. 
J. Voevodsky, Plato/Aristotle: A Neural Net Knowledge Processor, IEEE First International Conference on Neural Networks, XP002058597, Jun. 21 24, 1987, San Diego, CA, pp. 399 408. * 
L.C. Shiue and R.O. Grondin, "On Designing Fuzzy Learning NeuralAutomata," IEEE First International Conference on Neural Networks, XP002058598, Jun. 2124, 1987, San Diego, CA, pp. II299 through 11307. 
L.C. Shiue and R.O. Grondin, On Designing Fuzzy Learning Neural Automata, IEEE First International Conference on Neural Networks, XP002058598, Jun. 21 24, 1987, San Diego, CA, pp. II 299 through 11 307. * 
Lippmann, R. P., "An Intro. to Computing with Neural Nets," IEEE ASSP Mag. Apr. 1987, 422. 
Lippmann, R. P., An Intro. to Computing with Neural Nets, IEEE ASSP Mag. Apr. 1987, 4 22. * 
Procyk et al., "A Linguistic SelfOrganizing Controller", Automatica, v.15, pp. 1530 Jan. 1979. 
Procyk et al., A Linguistic Self Organizing Controller , Automatica , v.15, pp. 15 30 Jan. 1979. * 
Revised GMDH Algorithm Estimating Degree of the Complete Polynomial , pp. 928 934, 1986. * 
Takagi et al., "Artificial Neural Network Driven Fuzzy Reasoning" Intl. Workshop on Fuzzy System Applications, Aug. 1988, 217218. 
Takagi et al., Artificial Neural Network Driven Fuzzy Reasoning Intl. Workshop on Fuzzy System Applications, Aug. 1988, 217 218. * 
Cited By (4)
Publication number  Priority date  Publication date  Assignee  Title 

US6338051B1 (en) *  19980430  20020108  Samsung Electronics Co., Ltd.  User preference modeling method using fuzzy networks 
US6654763B2 (en)  20010614  20031125  International Business Machines Corporation  Selecting a function for use in detecting an exception in multidimensional data 
US20070260357A1 (en) *  20040609  20071108  Jorg Issberner  Method for the Production of Hydrophilic Polymers and Finishing Products Containing the Same Using a ComputerGenerated Model 
US20150331921A1 (en) *  20130123  20151119  Hitachi, Ltd.  Simulation system and simulation method 
Similar Documents
Publication  Publication Date  Title 

Hush et al.  Progress in supervised neural networks  
Kasabov et al.  FuNN/2—a fuzzy neural network architecture for adaptive learning and knowledge acquisition  
Juang  A TSKtype recurrent fuzzy network for dynamic systems processing by neural network and genetic algorithms  
Cheng  A multilayer neural network model for detecting changes in the process mean  
Stylios et al.  Fuzzy cognitive maps in modeling supervisory control systems  
Linkens et al.  Learning systems in intelligent control: an appraisal of fuzzy, neural and genetic algorithm control applications  
Setiono  A penaltyfunction approach for pruning feedforward neural networks  
Fahlman et al.  The cascadecorrelation learning architecture  
US4730259A (en)  Matrix controlled expert system producible from examples  
US5546503A (en)  Apparatus for configuring neural network and pattern recognition apparatus using neural network  
Lin et al.  Reinforcement learning for an ARTbased fuzzy adaptive learning control network  
Foster et al.  Neural network forecasting of short, noisy time series  
Chen et al.  A feedforward neural network with function shape autotuning  
De Veaux et al.  A comparison of two nonparametric estimation schemes: MARS and neural networks  
Blanco et al.  A realcoded genetic algorithm for training recurrent neural networks  
Gupta et al.  On fuzzy neuron models  
Reed  Pruning algorithmsa survey  
Lin et al.  Neuralnetworkbased fuzzy logic control and decision system  
Papageorgiou et al.  Active Hebbian learning algorithm to train fuzzy cognitive maps  
US5796611A (en)  Weather forecast apparatus and method based on recognition of echo patterns of radar images  
Castro  Fuzzy logic controllers are universal approximators  
Azeem et al.  Generalization of adaptive neurofuzzy inference systems  
US5255347A (en)  Neural network with learning function  
US5214746A (en)  Method and apparatus for training a neural network using evolutionary programming  
Morita  Associative memory with nonmonotone dynamics 
Legal Events
Date  Code  Title  Description 

FPAY  Fee payment 
Year of fee payment: 8 

FPAY  Fee payment 
Year of fee payment: 12 