JPH0679284A - Pre/middle chlorination dosage controller for purification plant - Google Patents

Abstract

PURPOSE:To enable automatic operation with an optimum dosage in spite of an abrupt fluctua tion in the weather and water quality by inferring a prechlorine dosage dividedly to the terms based on predicted meteorological factors and water quality factors, further inferring a correction rate and combing these values to determine a target prechlorine dosage. CONSTITUTION:Respective inferring means 101 to 104 for inferring a predicted insolation quantity from the predicted weather 1 and the predicted climatic temp. 2, the meteorological factor terms from the predicted insolation quantity 4 and the predicted water temp. 3, a predicted chlorine demand from the predicted org. matter concn. 5 and predicted inorg. matter concn. 6 of raw water and the water quality factors of the prechlorine dosage from the predicted chlorine demand and the predicted water temp. are provided. The insolation quantity correction term of the prechlorine dosage from the difference between the predicted and the present insolation quantity, the chlorine demand correction term of the prechlorine dosage from the difference between the predicted and the present chlorine demand 7, the org. matter concn. correction term from the target and the present org. matter concn. 8, 9 of the outflow of a treatment basin and the residucal chlorine concn. correction term from the target and the present residucal chlorine concn. 10, 11 are inferred by respective inferring means 105 to 108. The target prechlorine dosage 12 is determined by summing up the respective terms of the prechlorine dosage by an arithmetic means 109.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION This invention is used in
Middle chlorine injection control device

[0002]

2. Description of the Related Art In a water purification plant for water supply treatment, raw water such as river water is temporarily stored in a landing well 1201 as shown in FIG. 12, where pre-chlorination is carried out, followed by rapid agitation basin 1202 and flock formation basin. Iron, manganese, algae, etc. in the raw water are removed by passing through 1203 and the sedimentation tank 1204.
Next, medium chlorine is injected and then passed through the filter basin 1205, followed by post chlorine injection, and then sent through the clean water basin 1206 and distributed. In the figure, 1207 is a flash mixer, and 120
Reference numeral 8 is a flocculator paddle, and 1209 is an inclined plate settling device.

Here, the purpose of pre-chlorine injection is to remove ammonia nitrogen (NH ₃ —N), total iron (Fe), total manganese (Mn), and algae treatment in raw water. Due to deterioration, the number of water treatment plants that perform pre-chlorine injection treatment is increasing. The actual pre-chlorination process at the water treatment plant is
Water quality factors of raw water (water to be treated) such as NH ₃ -N concentration, K
The pre-chlorine injection rate is artificially determined by considering MnO ₄ consumption, total iron concentration or total manganese concentration, and also considering meteorological factors such as season, weather, insolation or wind speed to determine the water quality and weather. The actual situation is to manually change the chlorine injection rate before each change.

[0004]

The water quality factors that determine the pre-chlorine injection rate are measured for NH ₃ -N concentration, KMnO ₄ consumption, chlorine demand, total iron, total Mn concentration, etc. Since it is discontinuous data (once a day) obtained by manual analysis, there is a problem such as control delay when performing continuous injection processing. Also, the water quality of the day is a prediction from the water quality analysis results of the previous day, so an error will occur if the water quality of the day greatly changes from the previous day.

On the other hand, with respect to the meteorological factor as well, the uncertain factor is large because the expected temperature, temperature, solar radiation amount, etc. of the day are used. Therefore, if the actual weather deviates significantly from the expected weather, the pre-chlorine injection rate must be artificially changed.

[0006] In addition, before the chlorine is injected into the landing well and the result is known, it is 3 to 4 which corresponds to the residence time of the facility.
It takes about time, and therefore, the injection rate must be determined in consideration of various variations during that time. If this variation is not taken into consideration, the residual chlorine in the precipitated water after 3 to 4 hours will not be constant and will vary greatly. The operator of the water purification plant decides the pre-chlorine injection rate by artificially and empirically considering these factors, but even if it exceeds the expected fluctuation, insufficient injection or excessive injection will occur and the prescribed treatment effect will occur. Is not obtained, and there is a possibility that the water quality standard (trihalomethane amount 0.1 mg / l) will be exceeded. Further, it is a great burden on the operator to artificially change the pre-chlorine injection rate to approach the optimum value while judging various information, and the number of times is usually about 5 to 10 times a day. is there.

The present invention has been made under such circumstances, and an object thereof is to rapidly change factors of weather and water quality in a device for controlling pre-chlorine injection or medium-chlorine injection control of a water purification plant. It is to provide a device that can automatically operate at an optimum injection rate.

[0008]

The structure of the present invention is shown in FIG. 1 or FIG. The pre-chlorine injection control device for a water purification plant according to the present invention is a water purification plant that introduces raw water into a treatment basin consisting of a floc formation pond and a sedimentation basin. Be prepared.

(1) A means 101 for inferring an expected amount of solar radiation from expected weather and expected temperature.

(2) Means 102 for inferring the meteorological factor term of the pre-chlorine injection rate from the predicted solar radiation amount and the predicted water temperature.

(3) Means 103 for inferring the expected chlorine demand from the expected organic matter concentration and the expected inorganic matter concentration of the raw water.

(4) A means 104 for inferring the water quality factor term of the pre-chlorine injection rate from the predicted chlorine demand and the predicted water temperature.

(5) Means 105 for inferring a solar radiation amount correction term of the pre-chlorine injection rate from the difference between the estimated solar radiation amount and the present solar radiation amount.

(6) Means 106 for inferring the chlorine demand correction term of the previous chlorine injection rate from the difference between the predicted chlorine demand and the present chlorine demand.

(7) A means 107 for inferring an organic matter concentration correction term for the pre-chlorine injection rate from the difference between the target organic matter concentration and the present organic matter concentration of the treated pond effluent.

(8) Means 108 for inferring the residual chlorine concentration correction term of the previous chlorine injection rate from the difference between the target residual chlorine concentration and the present residual chlorine concentration of the treated pond effluent.

(9) Means 109 for obtaining the target pre-chlorine injection rate by summing up each term of the pre-chlorine injection rate.

Further, the medium chlorine injection control device of the water purification plant according to the present invention is a medium water chlorine injection control device in a water purification plant in which a filtration basin is arranged at the rear stage of a treatment basin consisting of a floc formation basin and a sedimentation basin. A device for controlling injection is provided with the following means.

(1) A means 201 for inferring an expected residence time in the filter from the amount of water in the filter.

(2) A means 202 for inferring the meteorological factor term of the medium chlorine injection rate from the expected residence time, the present amount of solar radiation and the present temperature element of the filter basin.

(3) Means 2 for inferring the expected chlorine demand from the current organic matter concentration of the inflow water of the filter and the present chlorine demand
03. The means 203 can make an inference with the presence or absence of manganese removal treatment in the filter basin as an input parameter in addition to the current organic matter concentration and the current chlorine demand.

(4) Means 204 for inferring the water quality factor term of the medium chlorine injection rate from the expected chlorine demand and the current temperature element of the filter basin.

(5) Means 205 for inferring a solar radiation amount correction term for the medium chlorine injection rate based on the fluctuation of the solar radiation amount.

(6) Means 206 for inferring an organic matter concentration correction term for the medium chlorine injection rate from the difference between the target organic matter concentration and the present organic matter concentration of the effluent of the filtration basin.

(7) Means 207 for inferring the residual chlorine concentration correction term of the medium chlorine injection rate from the difference between the target residual chlorine concentration and the present residual chlorine concentration of the effluent of the filtration pond.

(8) Means 208 for obtaining the target medium chlorine injection rate by summing up each term of the medium chlorine injection rate.

Here, in the above-mentioned front / medium chlorine injection control device, it is preferable that the present organic matter concentration used for various inferences is one detected by using the ultraviolet absorbance as an index. Moreover, it is preferable to use fuzzy inference for inferring various parameters. Note that in obtaining the difference between the input parameters in the inference of various correction terms, not only the deviation of the input parameter may be used, but also the behavior over time (for example, the rate of change of the deviation) may be used to make the inference.

[0028]

[Operation] In the above-mentioned chlorine injection control device, before the control,
The predicted solar radiation amount and the weather factor term of the injection rate are deduced from the predicted weather factors such as the predicted weather and the predicted temperature, and the predicted solar radiation pattern of the control day, for example, is obtained. Also, from the water quality factors such as the predicted organic matter concentration and the predicted inorganic matter concentration of raw water, the water quality factor term of the predicted chlorine demand and the injection rate is obtained.

Further, at the time of control, the pre-target chlorine injection rate is obtained by summing the above-mentioned meteorological factor term and water quality factor term. At this time, since the meteorological factor and the water quality factor are feedforward factors, a correction term is obtained based on various feedback factors and used to calculate the pre-target chlorine injection rate. Feedback factors include deviations between the current values of predicted weather factors and predicted water quality factors, deviations between the target organic matter concentration and the current organic matter concentration of the treated pond effluent, deviations between the target residual chlorine concentration and the current residual chlorine concentration, etc. To use.

Also in the medium chlorine injection control, the target medium chlorine injection rate is similarly obtained by using the meteorological factor / water quality factor as a feedforward factor and various correction terms as feedback factors. However, in this case, since the retention time is relatively short in inferring the meteorological factor term of the injection rate, the estimated retention time in the filter is inferred and the meteorological factor term of the infusion rate is inferred from this estimated retention time and the current meteorological factor. On the other hand, the method of detecting the variation of the amount of solar radiation and obtaining the correction term of the meteorological factor term based on this variation is adopted.

[0031]

Embodiments of the present invention will be described below. Figure 3
FIG. 3 shows a basic hardware configuration of a pre-chlorine injection rate control system according to an embodiment of the present invention. In the figure, 301
Is a process control device for controlling various processes, and 302 is a fuzzy arithmetic device, which includes a fuzzy inference section (inference engine) and a knowledge database. Reference numeral 303 is a central processing unit that performs main control of the present apparatus, 304 is an input device, and 305 is a display device. Further, reference numeral 306 is a group of instruments for measuring chlorine concentration, organic matter concentration, etc. at the outflow stage of the sedimentation tank. In this embodiment, a meter using ultraviolet absorbance as an index is used as the organic matter concentration meter. An injection device 307 injects pre-chlorine into raw water.

FIG. 4 shows a basic configuration of a control system realized by the process control device 301, the fuzzy arithmetic device 302, the central processing unit 303 and the like. Reference numeral 401 denotes a fuzzy inference unit, which performs fuzzy inference according to an algorithm to be described later, and each term D ₁ , D ₂ , ΔD ₁ ,
It outputs ΔD ₂ , ΔD _a , and ΔD _b . The control rules and membership functions used in the fuzzy inference unit 401 are stored in the knowledge database storage unit 402 in advance. Examples of these control rules and membership functions are shown in FIGS. Man-machine interface 40
Reference numeral 3 is for inputting settings such as fuzzy rules, and is provided with an input device such as a keyboard and a mouse. The manual input unit 404 captures the input expected weather, PAC injection rate, and the like. The process data input unit 405 takes in the residual chlorine concentration and the UV (ultraviolet absorbance) value, and inputs the input data used for fuzzy inference (for example, the residual salt deviation e and the residual salt deviation change rate de / dt for the residual chlorine concentration). To generate. These input data are input to the memory in the fuzzy inference unit 401 via the input data processing unit 408.

The inference result output processing unit 406 includes inference values D ₁ , D ₂ , ΔD ₁ , ΔD ₂ , ΔD _a , of the fuzzy inference unit 401.
The injection amount calculation unit 40 performs the output processing of ΔD _b.
In reference numeral 7, the inferred value is added together with the pre-chlorine injection rates D ₃ and D ₄ to obtain the target pre-chlorine injection rate and output to the injection device.

FIG. 7 shows an algorithm for each term of the pre-chlorine injection rate. The operation of the embodiment apparatus will be described with reference to this figure. First, before the control day, weather factors such as expected weather and water quality factors such as expected inorganic substances are set and input, and the pre-fuzzy inference is performed using these input data. Then, the chlorine injection rates D ₁ and D ₂ are inferred.

FIG. 8 shows the effect of solar radiation on chlorine consumption.
As can be seen from this figure, the amount of solar radiation has a great influence on the trend of chlorine consumption. Therefore, the expected insolation amount is inferred from the expected average temperature during the day on the day of control execution and the expected weather during the day, and the pre-chlorine injection rate D ₁ based on meteorological factors is calculated from the estimated insolation amount and the expected average water temperature during the day. Ask for. FIG. 9 shows the correlation between the weather and the amount of solar radiation. The estimated solar radiation amount to be inferred is the direct solar radiation amount or the total solar radiation amount during the time from sunrise to sunset. In the inference of the expected solar radiation amount, a fuzzy rule (inference rule) that uses the expected weather and the expected average temperature as the condition part (preceding part) and the expected solar radiation amount as the conclusion part (consequent part) is used. In the previous inference chlorine injection rate D _1, the expected amount of solar radiation and the expected average water temperature condition part and (antecedent), using the fuzzy rules which the pre chlorine injection rate D ₁ and Conclusions section (consequent). The numerical value of the forecast weather is assigned as 0 for rain or snow, 1 for cloudy weather, and 2 for clear weather.

Next, the pre-chlorine injection rate D ₂ determined from the water quality factor on the day of the control execution, for example, the required chlorine amount of inorganic matter, the required chlorine amount of organic matter and the average water temperature is obtained. In this inference, first, the expected chlorine demand is inferred from the expected inorganic substance (NH ₃ -N etc.) chlorine demand and the expected organic matter (KMnO ₄ consumption etc.) on the day of the control execution, and the expected chlorine demand and the expected average are calculated. The pre-chlorine injection rate D ₂ based on the water quality factor is inferred from the water temperature. The inference of the expected chlorine demand uses a fuzzy rule that uses the expected inorganic chlorine demand and the expected organic chlorine demand as the condition parts and the expected chlorine demand as the conclusion part. In the previous inference chlorine injection rate D _2, expected chlorine demand and the expected average water temperature of condition part, using fuzzy rules for a pre chlorine injection rate D ₂ and conclusion part. Since the reliability of the online measurement signal is low, it is desirable to execute the fuzzy inference based on the relational rules obtained from the expected inorganic substance amount and the expected organic substance amount as described above to calculate the chlorine demand amount.

The input parameters handled by the above inference are
Since predicted values are used for both weather and water quality factors, there is no chance of hitting with a probability of 100%. Therefore, in order to eliminate the error due to the difference between the actual value and the predicted value, the meteorological factor correction term ΔD ₁ and the water quality factor correction term ΔD ₂ are obtained, and the inferred values D ₁ and D ₂ are corrected. There is.

The meteorological factor correction term ΔD ₁ is used when the initially predicted weather is greatly deviated and the deviated state continues for a certain period of time, and the estimated solar radiation amount obtained in the first step portion 3 ₁ is used. It is inferred by the fuzzy rule that the current weather change is the condition part and the pre-chlorine correction injection rate ΔD ₁ is the conclusion part. As for the numerical display of the current weather change, a numerical value is assigned to each weather as described above, so the change amount of this numerical value is used. For example, if rain → clear, +2; if no change, ± 0; if clear → rain, −2. On the other hand, the water quality factors correction term [Delta] D ₂ is the predicted chlorine demand and condition part and a current estimated chlorine demand, chlorine correction injection rate [Delta] D ₂ prior corresponding to pre-correction amount of the chlorine injection rate D ₂ and conclusion part Inferred by fuzzy rules.

In some cases, the powder activated carbon injection treatment is carried out in the step before the landing well, and in this case, the adsorption amount of pre-chlorine by the powder activated carbon is added. If the pre-chlorine injection rate corresponding to this adsorption amount is D ₃ , D ₃ = k ₃ (= 1/5
~ 1/4) x powder activated carbon injection rate. Further, when algae and the like increase in the water purification facility, a pre-chlorine injection rate of about 5 to 10 ppm is added for algicidal treatment. The previous chlorine injection rate is D ₄ . These values are manually input if necessary.

Further, the concentration of organic matter in the effluent water is recognized using the UV value at the outflow stage of the settling tank as an index, and the pre-chlorine injection rate ΔD _a is determined according to the concentration. In this inference, a fuzzy rule is used in which the deviation e between the target value and the current value of the UV value and its change rate de / dt are used as a condition part, and the pre-chlorine injection rate ΔD _a is used as a conclusion part.

Further, according to the deviation between the target value of residual chlorine at the outlet of the sedimentation tank and the current residual chlorine amount, the pre-chlorine correction injection rate Δ
Find D _b . In this inference, the deviation (residual salt deviation) e between the current amount of residual chlorine and its target value and the temporal change rate de / dt of the residual salt deviation e are used as conditions, and the pre-chlorine correction injection rate Δ
A fuzzy rule whose conclusion is D _a is used.

If the solar radiation amount and the chlorine demand amount are predicted by fuzzy reasoning as in the above-mentioned embodiment, these sensors can be omitted, and in particular, the commercially available chlorine demand meter is unreliable and expensive, which is an advantage. Is big.

Here, when the pre-chlorine injection rate was manually set and the case where the pre-chlorine injection rate was determined by fuzzy inference as in the above-mentioned embodiment, the simulation operation was performed respectively,
The results shown in FIG. 10 were obtained. In the figure, (1) and (2) respectively show the injection rate and the residual chlorine amount when manual setting was performed, and (3) and (4) respectively show the injection rate and the residual chlorine amount when determined by fuzzy inference. . However, the chlorine injection amount using the fuzzy reasoning was 478 liters / day, and the chlorine injection amount using the manual setting was 474 liters / day. From the result of FIG. 10, it is understood that when the injection control is performed according to the present invention, the residual chlorine amount becomes closer to the target value than in the case of manual setting, and therefore the injection rate is appropriate.

Next, an example in which the present invention is applied to control of the medium chlorine injection rate will be shown. FIG. 11 shows the algorithm used to calculate the medium chlorine injection rate. In this case, since the residence time in the filter basin is relatively short, about 30 minutes, the medium chlorine injection rate D is calculated from the retention time and the current meteorological factors, instead of using the estimated solar radiation amount as in the calculation of the pre-chlorine injection rate. We will ask for ₁ . The retention time is inferred using a fuzzy rule with the amount of filtered water as the condition part and the retention time as the conclusion part. The medium chlorine injection rate D ₁ is inferred using a fuzzy rule in which the residence time, the amount of solar radiation, the current temperature, and the current water temperature are the condition parts, and D ₁ is the conclusion part.

The medium chlorine injection rate D ₂ due to the water quality factor is also inferred using the current value. That is, the estimated chlorine demand amount is obtained using a fuzzy rule in which the current UV value, the currently estimated chlorine demand amount, and the presence or absence of manganese removal treatment in the filter basin are used as the condition parts and the chlorine demand amount is used as the conclusion part. Furthermore, with the current water temperature and the estimated chlorine demand as the condition part,
The medium chlorine injection rate D ₂ is obtained by using a fuzzy rule whose conclusion is the medium chlorine injection rate D ₂ . If manganese removal treatment is performed in the filter basin, the load will be reduced and the chlorine consumption will be affected. Therefore, if necessary, the presence or absence of manganese removal treatment is manually set and used for inference.

Furthermore, the amount of solar radiation and U in the outflow stage of the filter basin
V value, the residual chlorine concentration of the feedback factor, the correction term [Delta] D ₁ of the middle chlorine injection rate, [Delta] D _2, obtains the [Delta] D _a. In the inference of the correction term ΔD ₁ , a fuzzy rule that uses the solar radiation deviation (change of solar radiation) e and its rate of change Δe / Δt as the condition part and ΔD ₁ as the conclusion part is used. In the inference of the correction term ΔD ₂ , a fuzzy rule is used in which the deviation e between the target value of the UV value and the measured current value and its change rate Δe / Δt are used as the condition parts, and ΔD ₂ is used as the conclusion part. In the inference of the correction term ΔD _a , a fuzzy rule is used in which the deviation e between the target value and the present measured value of the residual chlorine concentration and its change rate Δe / Δt are used as the condition parts, and ΔD _a is used as the conclusion part. Others are the same as the case of pre-chlorine injection rate control.

[0047]

As described above, according to the present invention,
In pre-chlorine injection control, the pre-chlorine injection rate is inferred by dividing it into terms based on the expected meteorological factors and terms based on the expected water quality factors, and is corrected according to the difference between those expected factors and actual factors. The amount of the chlorine is inferred, and the correction amount is inferred based on the organic matter concentration and the residual chlorine amount, and the pre-chlorine injection rate is determined by combining these values, so errors such as time delay can be suppressed as much as possible. In addition, it is possible to control the concentration of organic matter and the amount of residual chlorine at a constant level, and it is also possible to respond to sudden changes in weather and water quality factors.

Further, in the medium chlorine injection control, when the meteorological factor term of the medium chlorine injection rate is inferred, the residence time is inferred and the inference is performed based on this value and the present meteorological factor. It can be further suppressed.

Therefore, according to the present invention, the front / medium chlorine can be injected at the optimum injection rate, so that the injection amount can be significantly reduced as compared with the manual operation, and
The amount of trihalomethane can be reduced, and as a result, safe and delicious water can be supplied. Further, since the automatic calculation is performed, the burden on the operator can be reduced.

In addition to these effects, there are the following effects. Knowledge and experience of veteran operators can be expressed in the form of control rules and membership functions, and know-how can be shared. The control rules and membership functions can be improved and changed according to the situation, and the control algorithm can also be expected to improve. The inference result can be confirmed on a CRT or the like.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of the present invention.

FIG. 2 is a block diagram showing the configuration of the present invention.

FIG. 3 is an explanatory diagram showing a basic hardware configuration of the embodiment apparatus.

FIG. 4 is a block diagram showing an outline of a system configuration of the embodiment apparatus.

FIG. 5 is an explanatory diagram showing an example of a control rule.

FIG. 6 is an explanatory diagram showing an example of a membership function.

FIG. 7 is an explanatory diagram showing an algorithm of pre-chlorine injection rate calculation.

FIG. 8 is a graph showing the effect of solar radiation on chlorine consumption.

FIG. 9 is a graph showing a correlation between weather and the amount of solar radiation.

FIG. 10 is a graph showing the result of simulation operation.

FIG. 11 is an explanatory diagram showing an algorithm for calculating the medium chlorine injection rate.

FIG. 12 is a configuration diagram showing a water purification plant.

[Explanation of symbols]

 Reference numeral 101 ... Expected solar radiation amount inference means 102 ... Injection rate meteorological factor term inference means 103 ... Expected chlorine demand amount inference means 104 ... Injection rate water quality factor term inference means 105 ... Solar radiation amount correction term inference means 106 ... Water quality factor correction term inference means 107 ... organic matter concentration correction term inference means 108 ... residual chlorine concentration correction term inference means 109 ... pre-target chlorine injection rate calculation means

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Keiichi Tsukiashi 2-1-1-17 Osaki, Shinagawa-ku, Tokyo Inside the Meidensha Co., Ltd.

Claims (5)

[Claims] 1. A means for inferring an expected amount of solar radiation from an expected weather and an expected temperature in an apparatus for controlling pre-chlorine injection to the raw water at a water treatment plant introducing the raw water into a treatment pond including a floc formation pond and a sedimentation pond. And means for inferring the meteorological factor term of the pre-chlorine injection rate from the expected solar radiation amount and expected water temperature, means for inferring the expected chlorine demand amount from the expected organic matter concentration and expected inorganic matter concentration of the raw water, and the expected chlorine demand amount and Means to infer the water quality factor term of the pre-chlorine injection rate from the predicted water temperature, means to infer the solar radiation correction term of the pre-chlorine injection rate from the difference between the expected solar radiation amount and the present solar radiation amount, the expected chlorine demand amount and the present Means to infer the chlorine demand correction term of the pre-chlorine injection rate from the difference in chlorine demand, and the organic matter concentration of the pre-chlorine injection rate from the difference between the target organic matter concentration and the current organic matter concentration of the treatment pond effluent. The means for inferring the correction term, the means for inferring the residual chlorine concentration correction term for the pre-chlorine injection rate from the difference between the target residual chlorine concentration and the current residual chlorine concentration in the treated pond effluent, and the respective terms for the pre-chlorine injection rate are added And a pre-chlorine injection control device for a water purification plant, which is provided with means for obtaining a pre-target chlorine injection rate. 2. A device for controlling medium chlorine injection to the inflow water of a filter at a water treatment plant in which the filter is placed after the treatment tank consisting of a floc formation tank and a sedimentation tank, Means for inferring the expected residence time, means for inferring the meteorological factor term of the medium chlorine injection rate from the expected residence time, the current solar radiation and the current temperature element of the filter, the current organic matter concentration of the inflow water of the filter and the current chlorine Means to infer the expected chlorine demand from the demand, means to infer the water quality factor term of the medium chlorine injection rate from the estimated chlorine requirement and the current temperature element of the filter, and the medium chlorine injection rate based on the fluctuation of the solar radiation Means to infer the solar radiation correction term for the filter, the means to infer the organic matter concentration correction term for the medium chlorine injection rate from the difference between the target organic matter concentration and the current organic matter concentration in the effluent of the filter basin, and the target residue of the effluent in the filter effluent. A means for inferring the residual chlorine concentration correction term for the medium chlorine injection rate from the difference between the elemental concentration and the current residual chlorine concentration, and a means for obtaining the target medium chlorine injection rate by summing each term of the medium chlorine injection rate were provided. The chlorine injection control device in the water purification plant. 3. The apparatus according to claim 2, wherein the means for inferring the expected chlorine demand is inferred by using the presence or absence of manganese removal treatment in the filter basin as an input parameter in addition to the current organic matter concentration and the current chlorine demand. An inside chlorine injection control device for water purification plants characterized by being performed. 4. The device according to claim 1, wherein
A pre-chlorine injection control device or a medium chlorine injection control device of a water purification plant, wherein the present organic matter concentration is one detected using ultraviolet absorption as an index. 5. The device according to claim 1, wherein
A pre-chlorine injection control device or a medium-chlorine injection control device of a water purification plant, characterized by using a fuzzy inference means as a means for inferring various parameters.