JP2567968B2 - Automatic product inspection method for molding machines - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a product automatic inspection method for automatically determining the quality of a product in a molding machine such as an injection molding machine or a die casting machine.

[Prior Art] For example, when performing a molding operation by an injection molding machine in an automatic operation, it does not make sense that the molded product becomes a pile of defective products. The operating conditions are set in detail. Then, a microcomputer that controls the entire molding machine (hereinafter referred to as a microcomputer) executes automatic operation by referring to measurement data from various sensors based on preset molding operation condition values,
The molded product is designed to be molded continuously.

Also, together with the set value of the molding operation conditions described above, the upper limit value and the lower limit value are set in accordance with the respective set values,
Actually measuring how each molding operation condition value changed while performing automatic molding, and the measured value was set as above.
If it is within the range of the lower limit, it is judged as a good product, and if it is out of the set upper limit value or the set lower limit value, it is judged as a defective product. Recently, an injection molding machine with a so-called automatic inspection function, which is designed to be brought to a place other than the product accumulation (placement) place, has started to spread.

In the technology proposed by the applicant of the present application as Japanese Patent Laid-Open No. 3-36012, as an injection molding machine with an automatic inspection function, a microcomputer that controls the entire injection molding machine sets predetermined measurement data of each molding operation condition value. Capture the number of shots,
This is statistically calculated to determine the above-mentioned upper and lower limits.

By the way, in the technical field of injection molding, there are many unclear matters such as the details of the correlation of each molding operation condition and the detailed mechanism of the resin behavior, and research is underway to accurately grasp them. At present, the set values of the molding operation conditions and the upper and lower limit values set in the injection molding machine with the automatic inspection function depend on the setting by an operator who has abundant experience and knowledge. In view of this point, in the above-mentioned prior application, the upper and lower limit values can be set automatically by statistically calculating actual measurement data at the time of molding a good product.

[Problems to be Solved by the Invention] However, in the conventional product automatic discrimination method for an injection molding machine, all monitor items used for quality judgment correspond to the operating conditions of the injection molding machine. It does not include items such as molded product weight and molded product dimensions, which are closely related to molding quality, and the quality is judged before the molded product is opened and the molded product is taken out of the mold. "
It was a pass / fail judgment. For this reason, the allowable range tends to be set narrower in order to reliably exclude defective products, and it is pointed out that in the molded products excluded as defective products, molded products that are acceptable as good products in practice are mixed. was there.

In order to deal with this point, the applicant of the present application has filed Japanese Patent Publication No.
In the publication No. 16, a method was proposed in which the weight of the molded product was measured and taken in together with the actual measured value of each operating condition monitor item for each shot, and this weight of the molded product was used as one of the monitor items for the quality judgment. .

In the above-mentioned Japanese Patent Application No. 2-931, since the weight of the molded product is used for the quality judgment, the accuracy of the quality judgment is greatly improved, but on the other hand, since the weight of the molded product is measured for each shot, Weighing must always be performed after the mold opening / ejecting process, and it is necessary to operate the equipment for this weighing (molded product placing means, electronic scale, etc.) during continuous operation, which complicates operation control on the equipment side. Become.

At the time when the weighing is completed, a cycle for the next shot (for example, a charging operation) is started, and the process of associating the actual measurement value of the operating condition of each shot with the weighing information is troublesome.

Immediately after the mold opening / ejecting process (immediately after taking out the molded product from the mold), it is not possible to sort the good product and the defective product immediately, and it is necessary to sort the good / bad products after the measuring process. Side operation control becomes complicated.

There was a problem.

Therefore, the technical problem to be solved by the present invention is to solve the problems of the above-mentioned conventional technology, and the purpose thereof is to be an important factor for judging the quality of the product other than the operating condition monitor items. It is possible to apparently reflect a specific important monitor item, such as the weight of a molded product, in the good / defective judgment, and it is possible to automatically determine proper good / defective, and at the time of continuous automatic operation, the measured value of the specific important monitor item. An object of the present invention is to provide an automatic product inspection method for a molding machine that can predict and calculate data of a particular important item for each shot based on the actual measurement values of each operating condition monitor item without taking in.

[Means for Solving the Problems] In order to achieve the above-mentioned object, the present invention includes a microcomputer that drives and controls each unit of a molding machine based on set operating condition values and measurement information from each sensor, In order to determine the quality of the product during continuous automatic operation, the microcomputer has a function of taking in measured values of a plurality of predetermined operating condition monitor items and reflecting them in the product quality determination process. In the method, during a trial shot period of a predetermined number of shots, in addition to the operating condition monitor items, the measured values of specific important monitor items that are important factors for determining the quality of the product are loaded into the microcomputer, and the microcomputer Indicates the correlation between the specific important monitor item and each operating condition monitor item by the measured value of this specific important monitor item and the actual measured value of each other operating condition monitor item. The relational constant in the relational expression is calculated, and during continuous automatic operation, the microcomputer uses the correlation relational expression having the obtained relational constant without fetching the measured value of the specific important monitor item, The data of the specific important item is predicted / calculated for each shot based on the measured value of the monitor item, and the prediction / calculation result is used for the quality judgment of the product.

[Operation] For example, the microcomputer built in the injection molding machine is configured to monitor each operating condition such as an injection condition for each shot (each cycle) during a test shot period of a predetermined number of shots. In addition to the actual measurement values, the actual measurement values of the specific important monitor items, which are important factors for determining the quality of the product, are also captured in association with each shot, in addition to the operating condition monitor items. For example, if the specific important item is the weight of the molded product (product), the molded product taken out from the mold by the take-out machine for each shot during the trial shot period is measured by, for example, being mounted on an electronic scale.
The measured value of the weight of the molded product is sent from the electronic scale to the microcomputer.

Then, the microcomputer calculates a relational constant in the correlation equation between the molded product weight and each of the other operating condition monitor items based on the measured value of the molded product weight and the measured values of the other operating condition monitor items. That is, by the method of "multiple regression analysis by multivariate analysis method", the predicted value of the molded product weight is y ', the data of each operating condition monitor item is x ₁ , x ₂ , ... x _P , the measured value of the molded product weight. when was the _{y, y '= + b 1} (x 1 -x 1) + b 2 (x 2 - 2) + ...... + bP in represented by expressions (xP + xP), the constants b _1, b _2, ...... b _P is calculated (however, in the above equation, and ₁ , ₂ , ...
_P is the average value of each measured value). Each constant in the above equation is
It is calculated from the actual measured value by the strength of the degree of correlation between the data of each operating condition monitor item and the molded product weight data.
When the constant is determined, the predicted value y'of the weight of the molded product can be calculated in the Liam time at the time when the measured value of each operating condition for one shot is measured by the formula.

After the above-mentioned test shot, when the continuous automatic operation for continuously molding the molded product is started, the microcomputer captures the actual measurement value of each operating condition monitor item, and the actual measurement value and each preset monitor item In addition to comparing the lower limit value, predict and calculate the data of specific important monitor items such as molded product weight, and compare the predicted calculated value with the preset upper and lower limit values of the monitor item. Judgment is performed in real time for each shot until the mold is opened. Then, when it is determined that the shot is a good product, the molded product of the shot is carried to the regular product storage area, and when it is determined to be a defective product, the molded product of the shot is carried to a defective product accumulation position, for example, taken out. Control the machine.

By doing so, the quality determination can be performed in real time before opening the mold by a large number of monitor items including specific important items that are important factors for determining the quality of the product such as the weight of the molded product, and the accuracy of the quality determination is improved. . In addition, since it is not necessary to actually measure data other than operating condition monitor items such as molded product weight during continuous operation, but this is predicted and calculated from the actual measured values of each operating condition monitor item, the pass / fail judgment time is performed before the mold is opened. Thus, when the molded product is taken out from the mold, the good product and the defective product can be sorted.

[Examples] Hereinafter, one example shown in FIG. 1 and FIG. 2 in which the present invention is applied to an in-line screw type injection molding machine will be described. In the present embodiment, a hydraulic drive type injection molding machine will be described as an example, but the present invention can be similarly applied to a servomotor drive type injection molding machine.

FIG. 1 is an explanatory diagram showing a schematic configuration of a main part of an injection molding machine. The upper left part at the same time shows the mold opening / closing mechanism system. In the illustrated part, 1 is a base, 2 is a fixed die plate fixedly mounted on the base 1, and 3 is a base 1.
A support board installed on the slide base 1a extended to
Reference numeral 4 denotes a plurality of divers installed between the fixed die plate 2 and the support board 3. A mold clamping cylinder (hydraulic cylinder) 5 serving as a mold opening / closing drive source is fixedly mounted on the support board 3, and a known toggle link mechanism 6 is provided at the tip of the piston rod 5a of the mold clamping cylinder 5. Tie bar 4
The movable die plate 7 inserted in the is connected. Then, by moving the piston rod 5a forward and backward, the movable die plate 7 is moved toward or away from the fixed die plate 2.

Further, on the surfaces of the fixed die plate 2 and the movable die plate 7 which face each other, a fixed die 8 and a movable die 9 are provided.
And are installed. Then, during the mold closing process during the molding cycle, the toggle link mechanism 6 is extended by the advance of the piston rod 5a to advance the movable die plate 7, and the molds 8 and 9 are brought into close contact with each other. The toggle link mechanism 6 is stretched to provide a predetermined mold clamping force. On the other hand, during the mold opening process during the molding cycle, the piston rod 5a is retracted to collapse the toggle link mechanism 6 and the movable die plate 7 is retracted, so that both molds 8 and 9
Are separated from each other, and the molded product is taken out by a known eject mechanism (not shown) and an automatic take-out machine 10 for the molded product. Although not shown, the automatic take-out machine 10 includes, for example, a hand part for holding a molded product and an arm part for rotating and vertically moving the hand part, and is controlled by a microcomputer 30 described later.

In the present embodiment, the automatic take-out machine 10 described above,
When the molded product 50 taken out for each shot is judged as a non-defective product by comprehensively judging each monitor data as described later, the molded product 50 is placed on, for example, the belt conveyor 51, and the defective product is judged. If the item is
50 is put into the defective product storage 52. In addition,
During the trial shot period, the automatic take-out machine 10 is the total number of molded products.
50 is placed on the belt conveyor 50, and this trial shot is used by the electronic scale 11 with the communication function which is used only during the period,
The weight of the molded product is individually measured.

The upper right portion of FIG. 1 shows the injection mechanism system, in which 12 is a heating cylinder, 13 is a screw arranged in the heating cylinder 12 so as to be rotatable and forward and backward, and 14 is a heating cylinder 12. A nozzle attached to the tip, 15 is a band heater wound around the outer periphery of the heating cylinder 12, 16 is a hopper for supplying a resin material to the rear portion of the screw 13, and 17 is a motor (main drive) for rotating the screw 13. In the embodiment, for example, an electromagnetic motor is used,
It can be replaced with a hydraulic motor, etc.), 18 is a screw
It is an injection cylinder (hydraulic cylinder) for controlling the forward and backward movement of 13. As is known, the resin material supplied from the popper 16 is melted while being transferred to the tip side of the screw 13 while being kneaded and plasticized by the rotation of the screw 13, and the molten resin is stored at the tip side of the screw 13. The screw 13 is retracted while controlling the back pressure, and when one shot of molten resin is stored on the tip side of the screw 13, the screw rotation is stopped. Then, when the injection start timing is reached when a predetermined time has passed, the screw 13
Is driven forward to inject the molten resin into the cavity between the molds 8 and 9 which has been clamped.

20 is an injection pressure detection sensor composed of a hydraulic measuring head, etc.
Reference numeral 21 is an injection stroke detection sensor such as an encoder, 22 is a screw rotation detection sensor such as a rotary encoder, 23 is a temperature detection sensor that detects the temperature of the heating cylinder 12, and 24 is a molten resin temperature at the tip of the nozzle 14. A temperature detection sensor, 25 is a mold opening / closing stroke detection sensor including an encoder, 26 is a mold clamping pressure detection sensor including a hydraulic pressure measuring head, and 27 is an operation detection sensor of the automatic take-out machine 10. Measurement information signal
S1 ~ S8 and measurement information signals from other sensors not shown,
In addition to these measurement information signals during the test shot, the measurement information signal indicating the weight of the molded product from the electronic scale 11 is also added.
S9 is sent to the microcomputer 30, which will be described later, with appropriate input conversion processing as necessary.

30 is a microcomputer that controls the operation of the entire machine,
It controls the entire molding process such as mold opening / closing operation, charging operation and injection operation, and executes various arithmetic processing such as non-defective / defective product determination processing. The microcomputer 30 is actually a ROM that stores various I / O interfaces, main control programs and various fixed data, a RAM that reads and writes various flags and measurement data, a CPU (central processor unit) that controls the entire system, etc. Although various processing is executed in accordance with various programs created in advance, in the present embodiment, for convenience of explanation, the molding condition setting storage unit 31, the molding process control unit 32, the arithmetic processing unit 33, the measured value storage The following description will be made assuming that the unit 34, the upper / lower limit value setting storage unit 35, the comparison calculation unit 36, the constant calculation unit 37, the specific important monitor data calculation unit 38, and the like are provided.

The molding condition setting storage unit 31 stores various molding condition values input by the key input means 40 or other appropriate input means in a rewritable form after being subjected to arithmetic processing as necessary. The molding conditions include, for example, the relationship between the screw position, the screw rotation speed, and the back pressure during the charging process, suck back control conditions, and injection start point (position).
To the holding pressure switching point (position), the subdivided injection speed condition, the subdivided secondary injection pressure (holding pressure) condition from the holding pressure switching time to the holding pressure end time, the band heater temperature of each part, The mold closing stroke and speed, the mold clamping force, the mold opening stroke and speed, the eject control condition, the product takeout control condition and the like can be mentioned.

The molding process control unit 32 is based on a molding process control program created in advance and setting condition values stored in the molding condition setting storage unit 31, and is incorporated in the microcomputer 30 and the measurement information from the sensors 20 to 27. The corresponding drive source is drive-controlled through the driver group 41 while referring to the timing information from the generated clock to execute a series of molding steps. In FIG. 1, the drive signal D1 of the driver group 41 drives and controls the mold clamping cylinder 5 via the control valve 42, the drive signal D2 drives and controls the electric heat source of the band heater 15, and the drive signal D3 is The motor 17 is drive-controlled, the drive signal D4 drives and controls the injection cylinder 18 via the control valve 43, and the drive signal D5 drives the drive source (eg, motor, air cylinder, etc.) of the automatic take-out machine 10. In addition, other drive signals drive and control an appropriate drive source (not shown).

In the measured value storage section 34, all measured values x of preset monitor items are stored in the recording area over a predetermined number of consecutive shots. The monitor items that are taken in are roughly classified into time monitoring items, position monitoring items, rotation speed monitoring items, speed monitoring items, pressure monitoring items, temperature monitoring items, and power monitoring items. However, the factor closely related to the quality of the molded product is set as a monitor item in advance. Although the number of monitor items is arbitrary, the number of monitor items is 30 to 30 in this embodiment.
About 50, the measurement information from the sensors 20 to 27 and the time information from the clock built in the microcomputer 30 are converted as necessary and sequentially stored. The monitor items can be set by the operator by selectively inputting them.

In addition, in the present embodiment, in addition to the above monitor items, the molded product weight is set as a specific important monitor item, and during the trial shot period, measurement information indicating the molded product weight from the electronic scale 11 is displayed. Measured data by signal S9,
It is stored in the measured value storage unit 34. In addition, during the test shot period, it may be a case where molding of all non-defective products is guaranteed or a case where some defective products are mixed.

The arithmetic processing unit 33 statistically arithmetically processes the actual measurement value x for each operating condition monitor item when the data stored in the actual measurement value storage unit 34 reaches the predetermined number of sampling shots (that is, when the trial shot period ends). , The variation range R of the measured value x = (x _max −x _min ), the median value of the measured value x Me = (x _min + R / 2), and / or the average value of the measured value x = Σx _i ) / n standard deviation Then, the upper and lower limit values for each operating condition monitor item are set to the upper limit value = Me + a * R / 2 lower limit value = Me based on the above calculation result and the correction coefficient a preset by the experiential value. Calculated as −a · R / 2 or upper limit = + a · 3σ / 2 lower limit = −a · 3σ / 2. The upper and lower limits for each operating condition monitor item calculated in this way are transferred to and stored in the above-mentioned upper / lower limit setting storage unit 35. It should be noted that the upper and lower limits of each operating condition monitor item can be set to values that are much gentler than in the past, and the number of operating condition monitor items can be reduced. (This is because the weight of the molded product is used as a monitor item in this embodiment.) Similarly, the calculation processing unit 33 statistically calculates the measured value y of the molded product weight, which is a specific important monitor item, The upper limit value and the lower limit value of the weight of the molded product are calculated, and this is also transferred to and stored in the upper / lower limit value setting storage unit 35. For the automatic setting method of the upper and lower limits, refer to the above-mentioned prior application (Japanese Patent Application No. 1-169993) if necessary.

The constant calculating unit 37, the measured value y of the molded article parts during trial shot period, and x _1, x ₂ and the measured value of the operating condition monitoring items, when a ...... x _P, using these the method "multiple regression analysis by multivariate analysis"'shown in the aforementioned below represent, y' prediction value y of the molded article weight _{= + b 1 (x 1 -} 1) + b 2 (x 2 - 2) + …… + b _P (x _P − _P ), each constant b ₁ , b ₂ ,
…… Calculates b _P (However, in the above equation, and
₁ , ₂ , ... _P is the average value of each measured value). Each constant in the above equation is obtained by the degree of correlation between the data of each operating condition monitor item and the molded product weight data. The method of calculating the multiple regression formula is detailed in academic books and the like that represent various "multivariate analysis methods". Here, I will turn to the detailed scientific books, but recently, it has been marketed as "software for arithmetic calculation". Therefore, by using these, each of the above constants
It is relatively easy to calculate b ₁ , b ₂ , ... b _P.

The specific important monitor data calculation unit 38 includes a constant calculation unit 37.
And constants b _1, b ₂ ...... b _P obtained in, it found x ₁ for each operating condition monitoring items according to the state of the shot, x _2, using a ...... x _P, the predicted value of the molded article by weight by the formula Calculate y'in real time.

The comparison calculation unit 36 stores the data stored in the upper / lower limit value setting storage unit 35 and the actual measurement value data of each operating condition in the latest shot (for example, transferred from the actual measurement value storage unit 34).
And the predicted value y'of the molded product weight calculated by the specific important monitor data calculation unit 38, and compared with each measured value x ₁ , x ₂ , ...
... x _{P and} the predicted value y'of the weight of the molded product are within the upper and lower limit values (within the allowable range). And above
If it is out of the lower limit range, the molding process control unit 32 is made aware of this fact, and the molding process control unit 32 controls the driving of the automatic take-out machine 10. Is conveyed as a defective product to a predetermined defective product storage.

Here, in FIG. 1, 44 is a display device such as a color CRT display, 45 is a printer such as a dot printer, and the output results 44, 45 output the treatment results by the microcomputer 30 as necessary. To be done. Reference numeral 46 denotes an external memory such as a magnetic disk device, which exchanges information with the microcomputer 30 as needed.

In the present embodiment having the above-mentioned configuration, after the operation is started, it is confirmed that the good shots are continuously molded with stable shots by sampling the shots a predetermined number of times which is confirmed by weighing and visually observing the products. The above-mentioned upper and lower limit values for each operating condition monitor item for judging the quality of the product and the upper and lower limit values of the molded product weight (specific important monitor item) are set in the microcomputer 30. In addition, by sampling this predetermined number of shots, the above equation (multiple regression equation)
, Each constant b ₁ , b ₂ , ..., B _{P in} is calculated. After that, the microcomputer 30 determines the upper and lower limit values of each monitor item corresponding to the operating conditions and the upper and lower limit values of the molded product weight, and the corresponding actual measured value and molded product weight predicted / calculated value in the latest shot. By comparing each of the above items with each other, the above-described molded product good / defective determination processing and the sorting control of the molded product transport position based on the determination result are executed.

FIG. 2 is an explanatory diagram showing a part of the print output of the actual measurement value in a specific shot when the above-described molded article automatic inspection processing is executed, and includes monitor items, actual measurement values, set values, upper limit values, lower limit values, The good / bad determination marks are shown printed in the order of monitor items. (In practice, numerical values, unit display, and marks are printed on the blank part.) In the example shown in the figure, the monitor items are primary (injection) pressure, secondary pressure (holding pressure) switching position, 1 Next injection time, cushion position (screw forward most position), charge completion position (screw rearmost position), thermo (resin temperature of nozzle tip), charge time, cycle time, secondary pressure (holding pressure), ...
... The weight of the molded product (predicted calculation value) and the like are set.

As described above, in the present embodiment, a factor, which is closely related to the quality of the molded product, such as the weight of the molded product and the quality of the molded product, is included in the monitor item in real time for each shot (before the product is taken out). ) Since the quality is automatically discriminated, the discrimination accuracy is greatly improved. In addition, during continuous automatic operation (when executing the automatic product inspection process), the weight of the molded product is predicted and calculated by the multiple regression equation without actually measuring the weight of the molded product. It is not necessary to perform operation control for weighing or a process of fetching weighing data that is delayed, and when the molded product is taken out of the mold, the product can be sorted according to whether it is good or bad.

In the above-described embodiment, the weight of the molded product is mentioned as the specific important monitor item, but the specific important monitor item may also include the dimensions and appearance characteristics of the molded product (for example,
It is possible to adopt arbitrary items such as the degree of burr and sink marks, etc., and it is possible to further improve the discrimination accuracy by adopting a plurality of these items.

Further, in the above-described embodiment, each operating condition monitor item and the specific important monitor item are used for the good / defective determination, but in some cases, the good / defective judgment may be made only by the specific important monitor item. It is possible.

Needless to say, the present invention can be applied to a molding machine such as a die casting machine as well as the injection molding machine.

[Advantages of the Invention] As described above, according to the present invention, in addition to the operating condition monitor items, specific important monitor items that are important factors for product quality determination can be apparently reflected in the good / defective determination,
Appropriate good / defective automatic judgment is possible, and data of specific important item is recorded for each shot based on the actual measurement value of each operating condition monitor item without taking in the actual measurement value of specific important monitor item during continuous automatic operation. The seed molding machine has a great advantage that it can be predicted and calculated.

[Brief description of drawings]

1 and 2 relate to one embodiment of the present invention. FIG. 1 is an explanatory view showing a schematic configuration of a main part of an injection molding machine, and FIG. 2 is a specification when executing an automatic molded article inspection process. FIG. 8 is an explanatory diagram illustrating an example of print output of measured values in a shot. 1 ... Base, 2 ... Fixed die plate, 3 ... Support board, 4 ... Tie bar, 5 ... Mold clamping cylinder, 6 ... Toggle link mechanism, 7 ... Movable die plate, 8 ... Fixed side metal Mold, 9 ... Movable mold, 10 ... Automatic take-out machine, 11 ...
… Electronic balance, 12 …… Heating cylinder, 13 …… Screw, 14
...... Nozzle, 15 …… Band heater, 16 …… Hopper, 17
...... Motor, 18 …… Injection cylinder, 20 …… Injection pressure detection sensor, 21 …… Injection stroke detection sensor, 22 …… Screw rotation detection sensor, 23,24 …… Temperature detection sensor, 25
...... Mold opening / closing stroke detection sensor, 26 ...... Mold clamping pressure detection sensor, 27 …… Automatic take-out operation detection sensor, 30 ・ ・ ・
… Microcomputer, 31 …… Molding condition setting storage unit, 32 …… Molding process control unit, 33 …… Computation control unit, 34 …… Measured value storage unit, 35 …… Upper / lower limit value setting storage unit, 36 …… Comparison Arithmetic unit,
37: constant calculation unit, 38 ... specific important monitor data calculation unit, 40 ... key input means, 41 ... driver group, 42, 43 ...
… Control valve, 44… Display, 45… Printer, 46… External memory, 50… Molded product, 51… Belt conveyor, 52…
… A collection of defective products.

Claims (3)

(57) [Claims] 1. A microcomputer for driving and controlling each part of a molding machine based on each set operating condition value and measurement information from each sensor, the microcomputer comprising:
In order to determine the quality of the product during continuous automatic operation, a molding machine that has the function of importing the actual measured values of multiple predetermined operating condition monitor items and reflecting them in the product quality determination process During the number of trial shots,
In addition to the operating condition monitor item, the measured value of a specific important monitor item, which is an important factor of product quality judgment, is loaded into the microcomputer, and the microcomputer measures the measured value of the specific important monitor item and other operating conditions. Based on the measured value of the monitor item, the relational constant in the correlation equation between the specific important monitor item and each operating condition monitor item is calculated, and during continuous automatic operation, the measured value of the specific important monitor item is not taken in The computer predicts / calculates the data of the specific important item for each shot based on the measured values of the operating condition monitor items by using the correlation equation having the obtained relational constant, and calculates the prediction / calculation result. A method for automatically inspecting a molding machine product, which is used for judging the quality of a product. 2. The automatic product inspection method for a molding machine according to claim 1, wherein the specific important monitor items include at least product weight. 3. The automatic product inspection method for a molding machine according to claim 1, wherein the correlation equation is a regression equation by multiple regression analysis of a multivariate analysis method.