JPH04319326A - Cleaner - Google Patents

Abstract

PURPOSE:To minutely determine the sucking force of a fan motor and the rotating speed of a floor suction tool by the use of a neuro-fuzzy inferential equipment optimized by a learning rule from the detection value of a current sensor for detecting the current running to a fan motor. CONSTITUTION:A floor suction tool 4 for scraping up the dusts on a floor and a current sensor 1 for detecting the current running to a fan motor 2 are provided. The operation results obtained by operating the change ratio of the current and the absolute value of the current from the output of the current sensor 1 is subjected to the operation for optimizing various parameters of fuzzy operation according to learning rules such as fuzzy inferences based on membership function and re-diving method by a neuro-fuzzy inferential equipment 7, and the sucking force of the fan motor 2 and the rotating speed of the floor suction tool 4 are minutely determined.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vacuum cleaner that automatically adjusts suction force by detecting current flowing through a fan motor.

0002

2. Description of the Related Art Conventionally, in this type of vacuum cleaner, the suction power is set in about four levels depending on the amount of dirt, etc. In addition, there are some that set the suction force depending on the floor condition such as the length of the pile of the floor, tatami mat, or carpet.
Again, the condition of the floor surface could only be differentiated into three levels.

0003

[Problems to be Solved by the Invention] In general, with such conventional vacuum cleaners, the amount of garbage and the condition of the floor surface never reach 3 to 4.
It cannot be set in stages, but changes continuously, and the suction force of the fan motor also needs to be set in multiple stages, but since this cannot be accommodated, it is necessary to set the optimal suction power depending on the amount of clogged garbage and the condition of the floor surface. There is a problem in that the speed cannot be set, and at the same time, there is also a problem in that the rotation speed of the floor suction device cannot be controlled.

The present invention solves the above problems, and uses a neuro-fuzzy inference machine optimized by a learning rule from the detection value of a current sensor that detects the current flowing through the fan motor to determine the suction force of the fan motor and the floor pressure. The purpose is to precisely determine the rotation speed of the suction tool.

[0005]

[Means for Solving the Problems] In order to achieve the above object, the present invention includes a fan motor for sucking in dust, a floor suction device for scraping up dust on the floor, and detecting the current flowing through the fan motor. The neuro-fuzzy inference device is equipped with a current sensor, and a neuro-fuzzy inference device that optimizes various parameters of fuzzy inference using learning rules such as steepest descent method in order to determine the rotational speed of the fan motor and the floor suction device. The fuzzy reasoner is configured to determine the rotational speed of the fan motor and the floor suction device based on the output of the current sensor.

[0006]

[Operation] The present invention uses the above-mentioned problem-solving means to generate a neurotransmitter that is optimized by a learning rule from the output of a current sensor.
The suction force of the fan motor and the rotation speed of the floor suction device can be set using a fuzzy reasoner. Therefore, the suction force and the rotation speed of the floor suction device can be determined in detail, and the amount of dust in the dust collection chamber and the floor to be cleaned can be determined in detail. Dust can be removed efficiently regardless of the surface, and the operability can be greatly improved.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. 1 and 2.

As shown in the figure, a current sensor 1 is disposed near a power line 3 connected to a fan motor 2, and detects the current flowing through the fan motor 2. The current flowing to the fan motor 2 changes depending on the amount of air flowing into the fan motor 2 when the amount of dust in the dust collection chamber (not shown) increases or when the degree of close contact between the floor suction tool 4 and the floor increases. When decreases, the load on the fan decreases and the current decreases. The current sensor 1 detects the current flowing through the fan motor 2 that changes depending on the air volume, and detects the amount of dust in the dust collection chamber and the degree of contact between the floor suction tool 4 and the floor surface as a rate of change and absolute value of the current, It converts into an electrical signal, and is configured to obtain an analog output. The output of the current sensor 1 is calculated by the current change rate calculation means 5 and the current absolute value detection means 6. The neuro-fuzzy inference device 7 infers the suction force of the fan motor 2 and the rotation speed of the floor suction tool 4 from the outputs of the current change rate calculation means 5 and the current absolute value detection means 6. The control means 8 drives the fan motor 2 and the floor suction device 4 from the inferred values.

The neuro-fuzzy inference unit 7 is constructed as shown in FIG. 2, and the antecedent membership function storage means 9 stores membership functions relating to current change rate and current absolute value. Current change rate compatibility calculating means 10;
The current absolute value fitness calculation means 11 calculates the current change rate stored in the antecedent membership function storage means 9,
The degree of compatibility between the membership function regarding the absolute current value and the input current change rate and absolute current value is calculated. The antecedent minimum calculation means 12 includes the current change rate adaptation degree calculation means 10,
The MIN of the two compatibility degrees output from the current absolute value compatibility calculation means 11 is taken as the conclusion of the antecedent part. The suction force inference rule storage means 13 stores inference rules regarding suction force. The suction force membership function storage means 14 stores membership functions related to the suction force of the consequent part. The consequent part minimum calculation means 15 calculates the antecedent part conclusion and the suction force members of the consequent part stored in the suction force membership function storage means 14 according to the inference rules stored in the suction force inference rule storage means 13. The MIN of the ship function is taken as the conclusion of the rule. The center of gravity calculation means 16 obtains the respective conclusions for all the rules, takes the MAX of all the conclusions, and calculates the center of gravity to finally obtain the suction force. Further, a floor suction device rotation speed inference rule storage means 17 storing an inference rule for inferring the rotation speed of the floor suction device 4 and a membership function regarding the rotation speed of the floor suction device 4 are stored. Also included is a floor suction tool rotational speed membership function storage means 18.

[0010] This neuro-fuzzy inference device 7 can be easily realized by a microcomputer. Neuro-fuzzy inference device 7 includes antecedent membership function storage means 9, suction force inference rule storage means 13, suction force membership function storage means 14, floor suction device rotation speed inference rule storage means 17, and floor suction tool rotation speed inference rule storage means 17. The membership functions and inference rules stored in the suction device rotation speed membership function storage means 18 are the suction functions of the fan motor 2 that should be set in consideration of the current change rate, current absolute value data, and operational feeling during cleaning. Power and floor suction tool 4
is optimally set in advance using a learning rule such as the steepest descent method (one of the learning rules used in neural networks, which is a method of minimizing the error function) based on the rotation speed data. The control means 8 controls the fan motor 2 based on the determined suction force and the rotation speed of the floor suction device 4.
Then, the phase control amount of the floor suction tool 4 is calculated and controlled.

Next, to explain the operation of the above configuration, as dust is sucked in during normal cleaning and the dust in the dust collection chamber increases, the air volume flowing into the fan motor 2 decreases due to dust clogging, and the fan motor 2 As the load decreases, the current flowing tends to decrease. Therefore, by disposing a current sensor 1 near the power line 3 connected to the fan motor 2 and detecting changes in this current, the amount of dust in the dust collection chamber can be electrically detected. Furthermore, the output of the current sensor 1 changes depending on the floor surface to be cleaned. That is, when the object to be cleaned is a carpet, the floor suction tool 3 will stick to the carpet,
In the case of a wooden floor, the floor suction tool 4 hardly sticks to the wooden floor. Even in these two cases, there is a difference in the current flowing through the fan motor 2, and by detecting this as an absolute value of the current with the current sensor 1, it is possible to detect the difference in the floor surface to be cleaned.

Next, the process of inferring the suction force will be explained. The inference rule of the fuzzy inference in this example is ``If the amount of dust in the dust collection chamber is large (the current change rate is large) and the object to be cleaned is a carpet (the absolute value of the current is large), then the suction force is set to be very large. It is formed based on general judgments such as ``. Qualitative concepts such as the rate of change of current being "large", the absolute value of current being "small", and the suction force being "very large" are shown in Figures 3 (A) and (B) and Figure 4.
It is expressed quantitatively by membership functions as shown in (A) and (B). The current change rate compatibility calculation means 10 calculates the compatibility between the input from the current change rate calculation means 5 and the membership function regarding the current change rate stored in the antecedent membership function storage means 9 by MAX.
It is found by taking . Current absolute value suitability calculation means 1
1, the goodness of fit is determined in the same way regarding the input from the current absolute value detection means 6 and the membership function of the absolute value of the current stored in the antecedent membership function storage means 9. The antecedent minimum calculation means 12 takes the MIN of the two fitness degrees and uses it as the conclusion of the antecedent. The consequent minimum calculation means 15 calculates the antecedent conclusion and the suction force members of the consequent stored in the suction force membership function storage means 14 according to the rules stored in the suction force inference rule storage means 13. The MIN of the ship function is taken as the conclusion of the rule.

After determining the respective conclusions for all the rules, the center of gravity calculation means 16 takes the MAX of all the conclusions and calculates the center of gravity to finally determine the suction force. The control means 8 calculates and controls the phase control amount of the fan motor 2 based on the determined suction force. The rotation speed of the floor suction device 4 is determined by calculating the conclusion of the antecedent in the same manner as in the process of determining the suction force described above, and using the floor suction device rotation speed inference rule storage means 17 and the floor suction device rotation speed membership. The rotation speed of the floor suction tool 4 is determined from the function storage means 18.

[0014] In this embodiment, MAX is included in the inference method.
-The MIN synthesis method and center of gravity method are used, but other methods are also possible, and although the suction force, which is the consequent, is expressed using a membership function, it is also possible to express it using real values or linear form. Not even.

[0015]

[Effects of the Invention] As is clear from the above embodiments, according to the present invention, a fan motor for sucking in dust, a floor suction device for scraping up dust from the floor surface, and a current flowing through the fan motor are detected. comprising a current sensor and a neuro-fuzzy inference machine that optimizes various parameters of fuzzy inference using learning rules such as steepest descent method in order to determine the rotational speed of the fan motor and the floor suction device,
Since the neuro-fuzzy inference device determines the rotational speed of the fan motor and the floor suction device based on the output of the current sensor, it can finely determine the rotation speed of the fan motor and the floor suction device by fuzzy inference from the rate of change and absolute value of the current flowing through the fan motor. In addition, the optimal suction power and rotation speed of the floor suction device can be determined, allowing for efficient removal of dust regardless of the amount of dust in the dust collection chamber or the floor surface to be cleaned, and providing a vacuum cleaner that is extremely easy to operate. can do. Furthermore, as the number of inputs and outputs in fuzzy inference increases, it becomes difficult for humans to optimize the inference rules and their configurations between them, but the present invention uses learning rules such as steepest descent to・Since the configuration of the fuzzy inference machine is optimized, the above-mentioned effects can be obtained.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a vacuum cleaner according to an embodiment of the present invention.

[Figure 2
] Block diagram of the vacuum cleaner's neuro-fuzzy inference device

[Figure 3] (A) (B) Diagram showing the membership function of the same vacuum cleaner

[Figure 4] (A) (B) Diagram showing the membership function of the same vacuum cleaner

[Explanation of symbols]

1 Current sensor 2 Fan motor 4. Floor suction device 7 Neuro-fuzzy inference machine

Claims (3)

[Claims] 1. A fan motor for sucking dirt, a floor suction tool for scraping dirt from the floor surface, a current sensor for detecting a current flowing through the fan motor, and a rotation of the fan motor and the floor suction tool. a neuro-fuzzy inference machine that optimizes various parameters of fuzzy inference using a learning rule such as the steepest descent method to determine the A vacuum cleaner that determines the rotation speed of the floor suction device. [Claim 2] The neuro-fuzzy reasoner recognizes the amount of dust in the dust collection chamber and the degree of closeness between the floor suction device and the floor surface from the output of the current sensor, and determines the rotation speed of the fan motor and the floor suction device. The vacuum cleaner according to claim 1, wherein the vacuum cleaner is configured to: 3. The vacuum cleaner according to claim 1, wherein the neuro-fuzzy inference device optimizes the shapes of the antecedent membership function and the consequent membership function and the number of inference rules as various parameters.