JP4379238B2 - Coin identification device - Google Patents

Description

  The present invention relates to a coin discriminating device, and more particularly to a coin discriminating device having high fake coin elimination performance.

  Hereinafter, a conventional coin identifying device will be described. The conventional coin discriminating apparatus has a configuration as shown in FIG. In FIG. 8, reference numerals 1a, 1b, and 1c denote identification sensors for identifying the unevenness, material, and thickness of the inserted coins. The outputs of the identification sensors 1a, 1b, and 1c are connected to detection means 2a, 2b, and 2c that detect the characteristics of the coins and convert them into feature quantities representing the characteristics.

  The outputs of the detection means 2a, 2b, 2c are connected to one input of the comparison means 3a, 3b, 3c, respectively, and the output of the memory 4 is connected to the other input, respectively. The comparison means 3a , 3b, and 3c, the outputs of the detection means 2a, 2b, and 2c are compared with a determination range that is a reference for determination stored in the memory 4.

  The outputs of the comparison means 3a, 3b, 3c are connected to the determination means 5, and the correctness and denomination of the coins inserted by the determination means 5 are determined, and the determination result is output from the output terminal 6. Met.

  FIG. 9 is a characteristic diagram of output waveforms of the identification sensors 1a, 1b, and 1c. In FIG. 9, 7 is waveform data output from the unevenness sensor 1a, 8 is waveform data output from the material sensor 1b, and 9 is waveform data output from the material thickness sensor 1c. Here, the horizontal axis 10 is time, and the vertical axis 11 is the output level of the identification sensors 1a, 1b, 1c.

  As described above, the waveform data output from each of the identification sensors 1a, 1b, and 1c has a characteristic level and waveform according to the type of coin, and the difference in the output waveform causes the memory 4 to store the waveform data. Compared with the stored judgment range, the correctness of the coin and the denomination were identified.

As prior art document information related to the invention of this application, for example, Patent Document 1 is known.
JP 60-65393 A

  However, in such a conventional coin discriminating apparatus, since the determination range stored in the memory 4 is fixed, the fake coin exclusion capability and the denomination determination capability are limited. That is, the feature values output from the detection means connected to the identification sensors 1a, 1b, and 1c are distributed as shown in FIG. 10, for example. In FIG. 10, reference numeral 15 denotes a distribution of feature amounts output from the identification sensor 1 (any one of the identification sensors 1a, 1b, and 1c) when a true coin is inserted. This distribution 15 changes depending on external factors such as ambient temperature, and is not always constant. For example, the distribution 15 is a distribution as shown in FIG. Reference numeral 17 denotes a distribution of feature values output from the identification sensor 1 when a false coin (or a genuine coin of a different denomination) is inserted. As described above, the feature amount output from the identification sensor 1 has a range, and it has been difficult to clearly separate the correctness / incorrectness of the coin only from the fixed determination range. Since the denomination is completely the same, here, the denomination between the true coin and the fake coin is representatively described.

  In other words, in the feature amount output from the true coin and the fake coin, if the determination range for discriminating from the true coin is set as wide as 18, a portion 19 overlapping with the false coin is generated, and a coin corresponding to this portion 19 is inserted. If it is done, it will be misjudged. In order to discriminate between true coins and fake coins, a narrow range such as the judgment range 20 can reliably separate the true coins and fake coins, but the judgment range 20 regardless of whether it is a genuine coin. The part 21 which comes off is produced, and the specie within this part 21 must be eliminated. That is, there is a problem that the false coin elimination performance is deteriorated if the erroneous determination is eliminated.

  Here, the horizontal axis 23 is the frequency distribution of the feature amount, and the vertical axis 24 is the feature amount of the true or fake coin. There are two methods for storing the determination range 18 in the memory 4. One method is a method of storing the upper limit value 25 and the lower limit value 26 of the determination range 18 that is determined to be a genuine coin. The other method is a method of storing a reference value 27 corresponding to a characteristic amount of a true coin, and a positive-side allowable range 28a and a negative-side allowable range 28b with respect to this reference value. Both methods have the above-mentioned problems. The present invention solves such a problem, and an object of the present invention is to provide a coin discriminating apparatus capable of maintaining high discrimination performance even when external factors change.

  In order to achieve this object, the coin discriminating apparatus of the present invention is provided with a learning unit between the determination unit and the memory, and expands or contracts the determination range stored in the memory by the learning unit. Thereby, the initial purpose can be achieved.

  As described above, according to the present invention, the learning unit is provided between the determination unit and the memory, and the determination range stored in the memory is expanded and contracted by the learning unit.

  That is, with the above configuration, the determination range for determining the coins inserted by the learning means is automatically expanded and contracted. Therefore, since the acceptance range of the entered true coins is stabilized, the determination range of the true coins can be reduced, so that the false currency elimination performance can be improved while maintaining the acceptance performance of the true coins. .

  In addition, since the judgment range of the true coin can be automatically expanded in response to changes in external factors, even if the external factors change, it will follow these external factors while maintaining the acceptance performance of the true coins. Identification performance can be maintained.

(Embodiment 1)
FIG. 2 is a front view of the coin identifying device according to Embodiment 1 of the present invention. In FIG. 2, 121 is a coin slot, and 122 is a guide that is connected to the slot 121 and extends toward one vertical side surface 123 a of the casing 123. A snubber 124 formed of a material harder than any coin is provided on the extension of the guide 122. Reference numeral 125 denotes a passage connected to the snubber 124 and provided toward the other side surface 123 b of the housing 123.

  As shown in FIG. 3, an angle 126 of the passage 125 is tilted from 10 degrees to 12 degrees with respect to a horizontal plane, and coins roll naturally on the passage 125.

  Reference numerals 31 a, 31 b, and 31 c are identification sensors provided on the wall surface of the passage 125, and are arranged in this order from the insertion port 121. The first sensor 31a is located on the other side surface 123b from directly below the guide 122. Even if a forcible force is applied to the coin from the insertion port 121, the coin does not overlap with the sensor 31a and the snubber 124 or the passageway. At 125, the sensor 31a rolls at a substantially constant speed. Therefore, accurate recognition can be performed.

  Moreover, as shown in FIG. 3, the uneven sensor 31a which detects the unevenness | corrugation of a coin, the material sensor 31b which detects the material of a coin, and the material thickness sensor 31c which detects the thickness of a coin are located in this order. The centers of these identification sensors 31a, 31b, and 31c are provided at a fixed distance (13.25 mm) 130 from the passage 125.

  The diameter of the concave / convex sensor 31a is 8.3 mm, and the diameters of the material sensor 31b and the material thickness sensor 31c are both 12.5 mm. The distance 131 between the centers of the uneven sensor 31a and the material sensor 31b is 11.5 mm, and the distance 132 between the centers of the material sensor 31b and the material thickness sensor 31c is 13.5 mm. The distance 133 between the centers of the unevenness sensor 31a and the material thickness sensor 31b is set to 25 mm. By arranging the identification sensors 31a, 31b, and 31c in this way, the inserted coins always have a timing that is simultaneously detected by a plurality of identification sensors.

  Further, as shown in FIG. 4, the concave / convex sensor 31a and the material thickness sensor 31c are arranged such that two coils 134 and 135 are opposed to the passage 125, and these coils 134 and 135 have a negative mutual inductance. Series reverse phase connection. Therefore, since the unevenness and thickness of the coin 136 rolling in the passage 125 can be detected from both sides of the passage 125, the coin 136 is high regardless of the passage position of the coin 136 such as passing through the side wall of the passage 125 or the central portion of the passage 125. Detection accuracy is obtained. That is, the outputs of the coils 134 and 135 are affected by the passing position of the coin 136, and the respective outputs cause a decrease in detection accuracy. Is obtained.

  In the material sensor 31b, a sensor wound with coils 134 and 135 is arranged to face the passage 125, and these coils 134 and 135 are connected in series and in phase so that mutual inductance becomes positive. Therefore, since the material of the coin 136 can be detected from both sides of the passage 125, high detection accuracy can be obtained regardless of the passage position of the coin 136 such that the coin 136 passes through the side wall of the passage 125 or the central portion of the passage 125. That is, the outputs of the coils 134 and 135 are affected by the passage position of the coin 136 and cause a decrease in detection accuracy. However, when the detection is performed from both sides, the influence is offset, so that high detection accuracy is achieved. It is obtained.

  Returning to FIG. 2, the outputs of these identification sensors 31a, 31b, 31c are connected to a coin identification unit (including detection means 32a, 32b, 32c to output terminal 36 in FIG. 1 described later) 137, The identification unit 137 determines whether the coin 136 inserted is positive, false, or the coin 136 denomination (10 yen, 50 yen, 100 yen, 500 yen).

  Reference numeral 138 denotes a gate connected to the end of the passage 125, and the coin determined as “fake coin” by the identification unit 137 is guided to the outlet 140 through the back passage 139. Further, the coin 136 determined to be “positive” by the identification unit 137 is sorted into denominations by the sorting unit 141 and stored in the storage cylinders 142a to 142d for each denomination.

  Reference numeral 143 denotes a payout unit connected to the bottom side of the storage cylinder 142, and pays out a necessary number of coins 136 of a necessary denomination.

  FIG. 1 is a block diagram of a coin identifying device according to Embodiment 1 of the present invention. In FIG. 1, reference numerals 31a, 31b, and 31c denote identification sensors for identifying irregularities, materials, and material thicknesses of coins that are inserted. The outputs of the identification sensors 31a, 31b, and 31c are connected to detection means 32a, 32b, and 32c, and the outputs of these detection means 32a, 32b, and 32c are input to one input of the comparison means 33a, 33b, and 33c, respectively. The other input is connected to the output of the memory 34. The outputs of the comparison means 33a, 33b, 33c are connected to the determination means 35, and the output of the determination means 35 is connected to the output terminal 36.

  The detection means 32a, 32b, and 32c of the present embodiment each have an oscillation circuit composed of a capacitor and a positive feedback circuit connected to a coil constituting the identification sensor, and the oscillation waveform of the oscillation circuit is changed from a sine wave to an oscillation level. And a feature detecting means for detecting a maximum change amount from the oscillation level at the time of passing the coin and outputting it as a feature amount representing the feature of the coin.

  In addition, although the comparison means 33a, 33b, and 33c are provided in the output of the detection means 32a, 32b, and 32c, respectively, this is provided with an electronic changeover switch in the output of the detection means 32a, 32b, and 32c. If the switches are sequentially switched and supplied to the comparison means in a time-sharing manner, one comparison means is sufficient.

  A learning unit 37 is provided between the determination unit 35 and the memory 34. The learning means 37 has an input number counting section 38 connected to the output of the determination means 35, an input connected to the output of the insertion number counting section 38, and a calculation rate at which the output is connected to the memory 34. The determination unit 39 includes an insertion interval measurement unit 40 connected between the determination unit 35 and the calculation rate determination unit 39. A timer 41 is connected to the insertion interval measuring unit 40. Here, the comparison means 33a, 33b, 33c, the determination means 35, the learning means 37, the insertion number counting section 38, the calculation rate determination section 39, and the insertion interval measurement section 40 constituting the learning means 37 are software. It is configured.

  The operation | movement is demonstrated below about the coin identification device comprised as mentioned above. The coins inserted from the coin insertion port 121 pass through identification sensors 31 a, 31 b, and 31 c provided on the side surfaces of the passage 125 in the middle of rolling the passage 125 provided connected to the insertion port 121. . The identification sensor 31a mainly detects the unevenness of the coin, and the identification sensor 31b detects the material of the coin. The identification sensor 31c mainly detects the thickness of the coin.

  Coin characteristic data detected by these identification sensors 31a, 31b, and 31c is input to the detection means 32a, 32b, and 32c. In these detection means 32a, 32b, and 32c, the feature-value showing the characteristic of a coin is acquired from the input waveform data. The feature amount is compared with the determination range stored in the memory 34 by the comparison means 33a, 33b, and 33c. Based on the comparison result, the determination means 35 determines the correctness and denomination of the coins inserted. The determination result is output from the output terminal 36.

  FIG. 5 is an output waveform diagram of the identification sensor 31, and detection of the feature amount of coins will be described using this diagram. In FIG. 5, 61 is a waveform output from the unevenness sensor 31a, 62 is a waveform output from the material sensor 31b, and 63 is a waveform output from the material thickness sensor 31c. Here, the horizontal axis 64 is time, and the vertical axis 65 is the output level of the identification sensors 31a, 31b, 31c.

  In the present embodiment, the maximum change amount of the waveform output from each of the identification sensors 31a, 31b, and 31c is detected as the feature amount of the coin. That is, the maximum value 61c of the unevenness sensor 31a, the maximum value 62c of the material sensor 31b, and the maximum value 63c of the material thickness sensor 31c. Then, by comparing these feature amounts with the determination range, the inserted coins are determined.

  The learning unit 37 connected to the determination unit 35 counts the number of coins inserted by the insertion number counting unit 38 and outputs the result to the calculation rate determining unit 39. The calculation rate determination unit 39 calculates the calculation rate according to the number of coins inserted, and rewrites the coin determination range stored in the memory 34 based on the calculation result. The memory 34 stores a fixed initial determination range and a determination range that is output to the comparison means and used, and the latter is rewritten. Further, when the power is turned on, the latter is rewritten to the same value as the former and is reset to the initial determination range.

  Thus, the determination range compared by the comparison means 33a, 33b, 33c differs depending on external conditions such as input. That is, this determination range can be appropriately rewritten from the fixed initial determination range. The inserted number counting unit 38 is reset after the predetermined number is reached and the determination range is reduced.

  Further, the learning means 37 measures the insertion interval of coins inserted by the insertion interval measuring unit 40 using the timer 41. Then, the result is output to the calculation rate determination unit 39 and the timer 41 is reset. The calculation rate determining unit 39 changes the determination range by calculating the calculation rate according to the number of coins inserted and the insertion interval.

  Here, the operation of the learning means 37 will be described with reference to FIGS. The waveform signals output from the identification sensors 31a, 31b, and 31c are converted into feature quantities indicating the characteristics of coins by the detection means 32a, 32b, and 32c, and the distribution 44 shown in FIG. 6 is obtained. In FIG. 6, the horizontal axis 42 is a frequency distribution, and the vertical axis 43 is a feature amount of coins. This feature amount has a distribution having a spread with respect to the central value 45 due to the variation of each coin and the passage speed of the coins in the passage. In the present embodiment, the reference value 46a corresponding to the characteristic amount of the true coin and the allowable range 48a on the positive side and the allowable range 48b on the negative side with respect to this reference value are stored, whereby the determination range is stored in the memory. Stored. In addition, the calculation rate is stored for each denomination, and the product of the initial allowable range and the calculation rate is calculated to obtain an allowable range (hereinafter referred to as a practical allowable range) used in the comparison means. This calculation rate is stored in the memory 34 for each denomination and feature (unevenness, material, material thickness), and fine determination is possible.

  Reference numeral 47 represents a state when the power is turned on, and there is a deviation between the center value 45 and the reference value 46a of the distribution 44 of the true coins. This deviation is caused by, for example, a change over time from the time of product production until the time when a vending machine equipped with a coin identification device is installed in the market and starts operating. If the characteristic amount detected with respect to the inserted coin is within the determination range of the allowable range 48a on the positive side of the reference value 46a and the allowable range 48b on the negative side, it is determined to be a genuine coin.

  These allowable ranges 48a and 48b are rewritten based on the initial allowable range every time a correct coin is inserted, and a new practical allowable range is generated. In the present embodiment, the calculation rate is subtracted by 0.2% every time one genuine coin is inserted. In other words, the calculation rate determination unit 39 calculates the calculation rate by subtracting 0.2% each time a single coin is inserted. Accordingly, the determination range can be automatically reduced while maintaining the true coin acceptance performance, so that the false currency elimination performance can be improved. Note that the calculation rate has an upper limit of 100% and a lower limit of 70%, thereby preventing an unlimited expansion or reduction of the determination range.

  Further, the reference value 46a is automatically corrected every time a predetermined number of genuine coins are inserted in accordance with the insertion of the true coins. Reference numeral 49 in FIG. 6 represents a state after a certain number of genuine coins are inserted. The allowable range is reduced to 50a and 50b, and the reference value is 46b and the distribution 44 of the true coins. It approaches the center value 45. Further, when a true coin is inserted, as shown in 51, the allowable ranges 52a and 52b are minimized, and the reference value 46c matches the center value 45 of the distribution 44 of the true coins. Accordingly, it is possible to absorb a change with time and an error in the initial reference data.

  In FIG. 6, reference numeral 47 denotes a reference value 46a and determination ranges 48a and 48b as determination ranges stored in the memory 34 as initial values. The permissible ranges 48a and 48b are the same value in any coin discriminating apparatus, and cover the external environment such as the outside air temperature. Therefore, the range is large, and some data on the range of fake coins is included. In other words, the false currency rejection rate is bad.

  Therefore, as shown at 49, the allowable ranges 48a and 48b are sequentially narrowed to the allowable ranges 50a and 50b each time a genuine coin is inserted. Then, the newly determined value is stored in the memory 34 as a determination range. As a result, the unique identification of the coin identification device can be performed, and a determination range suitable for the external environment at the time of use can be obtained.

  Accordingly, since the determination range is reduced as the acceptance performance of the inserted true coins is stabilized, the fake coin elimination performance can be improved while maintaining the genuine currency acceptance performance.

  At this time, the reference value 46a is newly rewritten to the reference value 46b. Furthermore, if the number of times of insertion increases, it can be made narrower as in the determination ranges 52a and 52b shown in 51. The reference value 46c is also corrected. Here, the reduction speed of the allowable range is not more than the correction speed of the reference value.

  The permissible ranges 52a and 52b are lower limit values when the permissible range is reduced and reduced to the initial 70%. Therefore, the allowable range will not be further reduced no matter how much genuine coins are inserted thereafter.

  Here, if a plurality of initial allowable ranges are stored in the memory 34 and a hardware changeover switch for switching the allowable ranges is connected to the calculation rate determination unit 39, the calculation rate can be selected by the selection by the changeover switch. The calculation by the determination unit 39 is not performed. That is, the practical allowable range can be fixed to be the same as the initial allowable range. In that case, when operating with the switch between the normal tolerance range and the special emergency evacuation tolerance range when a fake coin is generated, the special tolerance range set narrower than the normal tolerance range is set. On the other hand, by not performing further reduction, it is possible to prevent harmful effects such as a decrease in the performance of accepting genuine coins.

  Next, expansion of the determination range will be described with reference to FIG. The distribution 44 of the true coins varies depending on external factors such as the ambient temperature, and is not always constant. For example, in some cases, the distribution is as shown in 57 or 59.

  In the learning means 37, when the time during which no coin is inserted is continued by the insertion interval measuring unit 40, the calculation rate, that is, the allowable ranges 56a and 56b is increased by 1.2% every 30 minutes. As a result, even if the external condition changes, for example, the environment such as the ambient temperature changes without any coins being inserted, it is possible to maintain the performance of accepting genuine coins. Therefore, according to the insertion interval of coins, it is possible to maintain a good balance between the true coin acceptance performance and the false coin elimination performance.

  Since the coin discriminating apparatus according to the present embodiment has the insertion interval measuring unit 40 in the learning means 37, the determination range of the discrimination data is automatically expanded by this insertion interval. In other words, if no coins are inserted for 30 minutes, the range is gradually increased by 30% every 30 minutes, such as the allowable ranges 56a and 56b, the allowable ranges 58a and 58b, and the allowable ranges 60a and 60b. When there is no input for a long time (12 hours in this embodiment), the discrimination data is returned to the original reference data. As a result, even if the surrounding environment changes, it is possible to maintain the performance of accepting genuine coins.

  In this embodiment, an example is given in which the determination range is reduced each time a correct coin is inserted, and the determination range is expanded every time, but the learning means accepts the coins that have been input so far. The same effect can be obtained by obtaining the rate and expanding / decreasing the determination range according to the acceptance rate. In this case, an accurate acceptance rate can be easily obtained by using the most recent predetermined acceptance rate.

  The coin discriminating apparatus according to the present invention is useful as a coin discriminating apparatus used in a vending machine or the like because it has a high false coin elimination performance.

The block diagram of the coin identification device in Embodiment 1 of this invention. Front view of the coin identification device Same part enlarged view Cross section of the passage Same as above, identification sensor waveform characteristics Schematic diagram showing reduction of judgment range Schematic diagram showing the expansion of the judgment range Block diagram of conventional coin identification device Same as above, output sensor waveform characteristics Same as above, output data distribution chart

Explanation of symbols

31a Concavity and convexity sensor 31b Material sensor 31c Material thickness sensor 32a Detection means 32b Detection means 32c Detection means 33a Comparison means 33b Comparison means 33c Comparison means 34 Memory 35 Determination means 36 Output terminal 37 Learning means 121 Input port 125 Passage

Claims (4)

A coin slot, a passage connected to the slot, an identification sensor disposed in the passage, a detection means connected to the identification sensor and detecting the characteristics of the coin, A memory storing a determination range as a denomination criterion for denominations, a comparison unit for comparing the determination range stored in the memory with a feature value output by the detection unit, and a coin according to a comparison result of the comparison unit Determining means for determining the authenticity of money and denominations, a learning means is provided between the determining means and the memory, the learning means expands and contracts the determination range stored in the memory, and at least in advance A coin identifying device that reduces the determination range each time a predetermined number of genuine coins are inserted . A coin slot, a passage connected to the slot, an identification sensor disposed in the passage, a detection means connected to the identification sensor and detecting the characteristics of the coin, A reference value corresponding to a characteristic amount of a genuine coin serving as a denomination criterion, and a memory in which a determination range is stored by storing a positive-side allowable range and a negative-side allowable range with respect to this reference value ; A comparison means for comparing the determination range stored in the memory with the feature value output by the detection means, and a determination means for determining the correctness of coins and the denomination based on the comparison result of the comparison means, A learning unit is provided between the determination unit and the memory, and the reference value stored in the memory is corrected each time a predetermined number of genuine coins are inserted by the learning unit, and the memory The size stored in Range is expanded and contracted, the speed to shrink at least the determination range, the coin discrimination apparatus according to the following compensation velocity of the reference value. The coin identifying device according to claim 1 or 2 , wherein the learning means expands the determination range every time a predetermined time elapses. The coin discriminating apparatus according to claim 1 or 2 , wherein an initial determination range and a calculation rate are stored in a memory, a determination range is defined by a product of these, and the calculation rate is automatically expanded and contracted.

Images