JP2011024325A - Method and device for averaging electricity-carrying load - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for averaging electricity-carrying loads which can average extreme increase and decrease of a power demand generated in a specified power time zone, and can average the loads in accordance with forms of a variety of power demands, and to provide a device. <P>SOLUTION: In the method for averaging the electricity-carrying loads when carrying electricity to a plurality of electricity-carried apparatuses in the specified power time zone from a time t1 to the time t1+a, the method is such that by calculating the latest electricity-carrying start time Tsm on the basis of an electricity-carrying time b of the electricity-carried apparatus, with respect to each electricity-carried apparatus by calculating an electricity-carrying adjustment coefficient k(0≤k≤1) by multiplying a random number rnd uniformly generated within a prescribed range by a preset random-number modified function fn, when determining an electricity-carrying time Ts of the electricity-carried apparatus from the electricity-carrying possible start time t1 to the electricity-carrying start time Tsm, and by determining the electricity-carrying start time Ts by the formula A: [Formula A] Ts=t1+k×(a-b). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

The present invention relates to a method and an apparatus for leveling an energized load of a device to be energized in a specific power time period such as a late-night power time period in which an electric charge is different from a normal time period. The present invention relates to an energized load leveling method and apparatus when energized equipment is energized in the late-night power hours.
In this specification, the midnight power from 11:00 pm to 7:00 am the next morning and the second midnight power from 1:00 am to 6:00 am may be collectively referred to as “midnight power”.
In this specification, “ordinary charging” refers to charging that is not rapid charging, and a secondary battery mounted on the vehicle is charged using a small charging device outside the vehicle having a 100V or 200V outlet. The “electric vehicle” refers to a vehicle whose power source is only a motor driven by a secondary battery, and a vehicle having a power source other than a motor driven by a secondary battery, such as an engine.

  Electricity companies will charge for late-night electricity charges with the goal of leveling power consumption (load leveling) by reducing the difference in power consumption between daytime and nighttime and between summer and other seasons. Various time-based contracts are available. However, when several thousand to several tens of thousands of electrical devices with a certain charge capacity start charging at the same time in a specific power time zone, the increase in power demand will increase significantly and then decline all at once. Problems arise. In order to solve such problems, various types of load leveling means using midnight power have already been put into practical use in electric water heaters that are widely used in homes (for example, Patent Documents 1 to 3). .

Moreover, in order to avoid starting and ending energization all at once, the electric power company provides individual energizable time setting timers (for example, Patent Document 4) for each region and performs energization control of the electric water heater. The operation is carried out so that the power load in the energized area at night is not concentrated.
However, since the number of the timers is very large, there is a problem that a lot of labor is required for management work such as periodically shifting the set value.

  Patent Document 5 discloses a method of allocating power demand in other time zones to power demand at the lowest load time in the midnight power time zone, and here, a method of adjusting the energization start time using a random number is proposed. Yes. However, Patent Document 5 is based on the premise that the required energization time is a certain time or less, and is not suitable for application to devices having a relatively long energization time, such as an electric water heater.

  Patent Document 6 proposes various methods for adjusting energization time using communication. However, when the number of load terminals is on the order of tens of thousands or more, this method is considered to be difficult to put into practical use because a problem of traffic countermeasures due to simultaneous communication occurs.

  By the way, in preparation for the spread of electric vehicles in recent years, it has been considered to provide a charging device that uses late-night power even in rental parking lots such as apartment buildings. At the same time, the problem of rapid increase in power consumption arises. At present, this type of problem does not occur because the penetration rate of electric vehicles is low. However, with the spread of electric vehicles in the future, the occurrence of power consumption peaks due to charging is considered to be a problem that cannot be ignored.

Japanese Patent Publication No. 3-18106 JP-A-6-180147 JP-A-6-180148 Japanese Patent Laid-Open No. 10-197069 Japanese Patent No. 4142837 Japanese Patent Publication No. 2008-160902

  The present invention provides an extreme increase / decrease in power demand that can occur when charging several thousand to several tens of thousands of electric devices (for example, electric water heaters and electric vehicles) having a constant charging capacity in a specific power time zone. The problem is to solve it by an autonomous mechanism.

  Also, the form of power demand for midnight power varies depending on the season and region, and it is difficult to adopt uniform load leveling means. Furthermore, in recent years, with the introduction of the second midnight power (from 1 o'clock to 6 o'clock) for the purpose of leveling off midnight power demand, not only a form having one negative peak but also a form having two negative peaks It is appearing (see FIG. 6). It is a problem to be solved by the present invention to achieve load leveling in accordance with such various forms of power demand.

  An object of the present invention is to provide an energization load leveling method and apparatus that can solve the above-described problems.

  In order to solve the above-mentioned problems, the inventor can invent load leveling autonomously by inventing the concept of a random number modification function and thereby controlling the form of power demand for a specific energized device. It was.

That is, the present invention relates to an energization load leveling method comprising the following technical means.
[1] An energization load leveling method for energizing a plurality of energized devices in a specific power time period starting from time t1 and ending at time t1 + a,
For each energized device, the latest energization start time Tsm is calculated based on the energization time b of the energized device, and the energization start time Ts of the energized device is determined between the energization possible start time t1 and the energization start time Tsm. In this case, the energization adjustment coefficient k (0 ≦ k ≦ 1) is calculated by multiplying the random number rnd uniformly generated in a predetermined range by a preset random number modification function fn, and the energization start time Ts is calculated by the following formula A. An electrified load leveling method characterized by determining.
[Formula A] Ts = t1 + k × (ab)
[2] The random number modification function fn is an inverse function of the cumulative distribution function P (k) of the probability density function p (k), the probability density function is defined by the following formula B, and the cumulative distribution function P (k) is The energization load leveling method according to the above [1], which is defined by Formula C.
[Formula B]
[Formula C]
[3] The random number modification function fn is one from a previously created cumulative distribution function group including a function of the following formula D, the following formula E, the following formula F, and / or the following formula G (where r is a positive integer). The electrification load leveling method according to the above [2], wherein a function is selected and set.
[Formula D] P (k) = 1
[Formula E] P (k) = (r + 1) × (1-k) ^ r
[Formula F] P (k) = (r + 1) × k ^ r
[Formula G] P (k) = (r + 1) × ((1-k) ^ r + k ^ r) / 2
[4] The function according to [3], wherein the function of the formula G is selected as the random number modification function fn when the specific power time zone is a midnight power time zone in which the second midnight power is set. Leveling method
[5] The energized load leveling method according to any one of [1] to [4], wherein the energized equipment includes an electric water heater and an electric vehicle.
[6] A program for causing a computer to execute the energization load leveling method according to any one of [1] to [5].

The present invention also relates to an energized load leveling device comprising the following technical means.
[7] An autonomous energized load leveling device that is electrically connected to an energized device that is energized in a specific power time period that starts at time t1 and ends at time t1 + a, and performs energization control,
The energization load leveling device calculates the latest energization start time Tsm based on the energization time b of the energized device, and sets the energization start time Ts of the energized device between the energization possible start time t1 and the energization start time Tsm. In the determination, the energization adjustment coefficient k is calculated by multiplying a random number modification function fn set in advance with a random number rnd that is uniformly generated within a predetermined range, and the energization start time Ts is determined by the following equation A. An energizing load leveling device.
[Formula A] Ts = t1 + k × (ab)
[8] The random number modification function fn is an inverse function of the cumulative distribution function P (k) of the probability density function p (k), the probability density function is defined by the following formula B, and the cumulative distribution function P (k) is The energization load leveling device according to [7], which is defined by Formula C.
[Formula B]
[Formula C]
[9] Storage means for storing a cumulative distribution function group including a function of the following formula D, the following formula E, the following formula F and / or the following formula G (where r is a positive integer), and the stored cumulative distribution The energization load leveling device according to [8] above, wherein one function selected from the function group is set as the random number modification function fn.
[Formula D] P (k) = 1
[Formula E] P (k) = (r + 1) × (1-k) ^ r
[Formula F] P (k) = (r + 1) × k ^ r
[Formula G] P (k) = (r + 1) × ((1-k) ^ r + k ^ r) / 2
[10] The above [7] to [7], wherein the function of the above equation G is set in the random number modification function fn when the specific power time zone is a midnight power time zone in which the second midnight power is set. 9] The energization load leveling device according to any one of [9].
[11] A charging device for an electric vehicle equipped with the energization load leveling device according to any one of [7] to [10].
[12] An electric water heater equipped with the current-carrying load leveling device according to any one of [7] to [10].
The “predetermined range” for the random number rnd in the above [1] and [7] is, for example, 0 ≦ rnd ≦ 1, 0 <rnd ≦ 1, 0 ≦ rnd <1 or 0 <rnd <1. It is.

ADVANTAGE OF THE INVENTION According to this invention, it becomes possible by an autonomous mechanism to eliminate the extreme increase / decrease in the electric power demand which may arise when charging electric devices, such as an electric water heater and an electric vehicle, in a specific electric power time slot | zone.
Moreover, by incorporating a simple load distribution means into the charging device of the electrical equipment, it becomes possible to realize load leveling according to various forms of power demand with an autonomous mechanism.

1 is a basic configuration diagram of an ordinary electric vehicle charging device equipped with load leveling means of the present invention. It is a basic lineblock diagram of an electric water heater carrying the load leveling means of the present invention. It is a conceptual diagram for demonstrating the setting of an energization start time. It is a graph for demonstrating the probability that it supplies with electricity. It is a graph for demonstrating a cumulative distribution function. It is a graph which shows the demand of the 2nd midnight electric power in a certain area. It is a calculation procedure of a graph of a front mountain type probability density function p (k) and a probability of energization Pon (t). It is a graph of the probability Pon (t) of energizing the front mountain when r = 5 and a = 8 hours. It is a calculation procedure of a graph of a back mountain type probability density function p (k) and a probability of energization Pon (t). It is a graph of the probability Pon (t) of energization of the rear mountain when r = 5 and a = 8 hours. It is a calculation procedure of a graph of the Futama type probability density function p (k) and the probability of energization Pon (t). It is a graph of the probability Pon (t) of energization of the double mountain type when r = 5 and a = 8 hours. It is a graph of the probability density function of a uniform type, a front mountain type, a rear mountain type, and a double mountain type. It is a graph of a cumulative distribution function of a uniform type, a front mountain type, a rear mountain type, and a double mountain type. It is a graph of a random type modification function of a uniform type, a front mountain type, a rear mountain type, and a double mountain type. It is function summary explanatory drawing of the load leveling means of this invention. It is an example of prediction of the charging time (normal charging) of an electric vehicle estimated from the daily travel distance of a passenger car. 3 is a graph of expected values of charging loads (normal charging) of 50 electric vehicles according to Example 1; It is a table | surface of the charging time of 50 electric vehicles which concern on Example 1, the pseudorandom number which generate | occur | produced, the electricity supply adjustment coefficient k, and charging start time. 6 is a graph of a composite load for each random number modification function according to the first embodiment. It is a graph of an expected value of energization load (normal charge) of 50 electric water heaters concerning Example 2. It is a graph of the expected value of the electricity supply load of 50 electric water heaters according to the second embodiment. It is a table | surface of the charging time of 50 electric water heaters based on Example 2, the generated pseudorandom number, the electricity supply adjustment coefficient k, and the charge start time. 10 is a graph of a composite load for each random number modification function according to the second embodiment.

  The mode for carrying out the present invention will be described with reference to << 1 >> electric vehicle ordinary charging device and << 2 >> electric water heater. The main variables that appear in the following description are:

<Main variables>
a: Energization time b: Energization time k: Energization adjustment coefficient m: Electric water heater adjustment coefficient t: Time within energization possible time t1: Energization possible start time t2: Energization possible end time Ts: Energization start time Te: Energization End time Tsm: Latest energization start time We: Combined power expected value

<< First Embodiment: Electric Vehicle Ordinary Charging Device >>
(1) Device Configuration FIG. 1 shows a basic configuration diagram of an ordinary electric vehicle charging device equipped with the load leveling means of the present invention. The apparatus includes an outlet that supplies electricity to an electric vehicle, a switch that opens and closes a circuit, a leakage breaker that protects the circuit from a short circuit and a ground fault, a power receiving device that receives electricity from an electric power company, and a control device that controls the switch Consists of
The battery level of the electric vehicle is sent from the electric vehicle by communication using a charging cable (PLC: power line communication), or if the electric vehicle does not have a communication function, the charger reads the battery level indicator of the electric vehicle. Is manually input to the input device. The control device calculates the energization time based on the remaining battery capacity, determines energization start / end times according to the energization adjustment logic, and turns the switch on and off at that time.
In addition, although the apparatus structure which can charge several units | sets simultaneously is shown in FIG. 1, the load leveling means of this invention is applicable also to the charging device for one unit | set. Although the power receiving capacity cannot be reduced by using only one charging device, from the viewpoint of the total charging demand in a certain region, it is possible to contribute to load leveling of the total charging demand by adjusting the charging start time.

(2) Determination method of energization start time
(2-1) Setting of energizable time zone “Energizable start time t1” and “energizable end time t2” are set.
In the midnight power time zone, t1 = 23: 00 and t2 = 7 o'clock (31:00) are set. In this case, the “energization possible time a” is 8 hours. Note that t1 and t2 can be changed for the convenience of the user (for example, charging is completed early in the morning).

(2-2) Calculation of energization time The “energization time b” required for charging the electric vehicle is calculated by Equation 1. In Equation 1, “electric vehicle rated charging time” is the time required for charging the remaining battery amount from 0% to 100%. In this specification, the term “energization” may be used to include charging.

[Expression 1] b = Electric vehicle rated charging time × (100% −battery fuel gauge indication value (%)) / 100

(2-3) Determination of energization start time
(2-3-1) Random number generation A random number between 0 and 1 is generated using the calculation function in the controller. The random number may be a pseudo-random number obtained by a squaring method, a mixed congruential method, or the like. In this specification, the term “random number” may be used in the meaning including pseudorandom numbers.

(2-3-2) Conversion of random numbers In order to adjust load distribution, random numbers with a fixed probability of occurrence follow a certain probability density function by multiplying the random numbers by a preset “random number modification function fn”. It is converted into a numerical sequence (energization adjustment coefficient k: 0 ≦ k ≦ 1). The calculation procedure of the energization adjustment coefficient k using the random number modification function fn will be described later.

(2-3-3) Determination of Energization Start Time and Energization End Time “Energization start time Ts” can be set to any time between the energization possible start time t1 and t1 + (ab). The energization start time Ts is obtained using Equation 2 using the energization adjustment coefficient k.
[Formula 2] Ts = t1 + k × (ab)

The “energization end time Te” can be obtained by adding the energization time b to the energization start time, as shown in Equation 3.
[Formula 3] Te = t1 + k × (ab) + b

<< Second Embodiment: Electric Water Heater >>
(1) Device Configuration FIG. 2 shows a basic configuration diagram of an electric water heater equipped with the load leveling means of the present invention. The apparatus includes an electric water heater body, a switch that opens and closes a circuit that supplies electricity to the electric water heater, an earth leakage circuit breaker that protects the circuit from a short circuit and a ground fault, a power receiving device that receives electricity from an electric power company, and a switch It consists of a control device to control.
The remaining amount of hot water in the electric water heater is estimated by temperature sensor data attached in the vertical direction of the electric water heater hot water storage tank, and the control device calculates the required energization time based on the remaining amount of hot water. The control device determines energization start / end times according to the energization adjustment logic, and turns the switch on and off at that time.
An electric water heater adopting the energization control system already has this configuration, and in that case, it can be dealt with only by changing the control software of the control device.

(2) Determination method of energization start time
(2-1) Setting of Energizable Time Zone The setting of “energizable start time t1” and “energizable end time t2” is the same as in the first embodiment. Further, in the case of the midnight power time zone, the “energization possible time a” is 8 hours.

(2-2) Calculation of energization time Various methods of calculating the required energization time of electric water heaters have been proposed, but basically, the average temperature of hot water in the hot water tank is obtained from multiple temperature sensors installed in the hot water tank, and the cooking is completed. The heat input required to raise the temperature to 85 ° C. can be obtained by dividing by the electric heater output, and this is defined as “energization time b” (see Equation 4).

[Formula 4] b = (85 ° C.−Hot water tank average temperature (° C.)) × Hot water tank volume (L) / (Electric heater output (kW) × 860 kcal / kW) × Electric water heater adjustment coefficient m
Electric water heater adjustment coefficient m: coefficient considering heat dissipation and container heat capacity (> 1)

(2-3) Determination of Energization Start Time and Energization End Time The determination of “energization start time Ts” and “energization end time Te” is the same as in the first embodiment (see the above formulas 2 and 3).

《Mathematical theory of load leveling using random numbers》
A mathematical theory of load leveling using random numbers will be described using an example in which a load in a specific energizable time zone such as a midnight power time zone (23:00 to 7 o'clock) is energized.
The relationship between “energization possible start time t1” and “energization possible end time t2” and “energization possible time a” is as shown in Equation 5.
[Formula 5] a = t2-t1

Assuming that the energization time is b (b ≦ a), the energization start time Ts can be selected at any time within the range from the energization possible start time t1 to the latest energization start time Tsm. “Latest energization start time Tsm” is calculated by Equation 6.
[Formula 6] Tsm = t2-b = t1 + ab

The energization start time Ts can be expressed by Equation 7 using an energization adjustment coefficient k (0 ≦ k ≦ 1).
[Expression 7] Ts = t1 + k × (ab)

A condition in which power is supplied at a certain time will be described with a specific example with reference to FIG.
At a certain time t1 + t (0 ≦ t ≦ a) between the energization start possible time t1 and the energization possible end time t2, the condition of energization is that t1 + t is within the energization time zone (ii in FIG. 3). Is represented by the following formula 8.
[Formula 8] t1 + k × (ab) ≦ t1 + t ≦ t1 + k × (ab) + b

Here, in the expression 8, since t1 is added to all sides, the following expression 9 is derived.
[Expression 9] k × (ab) ≦ t ≦ k × (ab) + b

Furthermore, the following formulas 9-1 and 9-2 are derived from the formula 9 for the energization adjustment coefficient k.
[Formula 9-1] k ≦ t / (ab)
[Formula 9-2] (tb) / (ab) ≦ k

  Here, it can be seen from FIG. 3 that the condition calculation formulas need to be different for b ≧ 0.5a (upper stage) and for b <0.5a (lower stage). That is, when b ≧ 0.5a, energization is always performed between a−b ≦ t ≦ b (the upper shaded portion in FIG. 3). In other words, it is understood that it is necessary to obtain a condition for energizing at time t1 + t.

(I) When b ≧ 0.5a (i−1) When 0 ≦ t <a−b, the left side of Equation 9-2 is negative, and thus the following Equation 10-1 is derived.
[Formula 10-1] 0 ≦ k ≦ t / (ab)

(I-2) When ab ≦ t <b, the right side of the equation 9-1 is 1 or more and the left side of the equation 9-2 is 0 or less, so the following equation 10-2 is derived. When t is within this range, energization is performed in all cases.
[Formula 10-2] 0 ≦ k ≦ 1

(I-3) In the case of b ≦ t ≦ a, since the right side of the equation 9-1 is 1 or more, the following equation 10-3 is derived.
[Formula 10-3] (tb) / (ab) ≦ k ≦ 1

(Ii) In the case of b <0.5a (ii-1) In the case of 0 ≦ t <b, since the left side of Expression 9-2 is negative, the following Expression 11-1 is derived.
[Formula 11-1] 0 ≦ k ≦ t / (ab)

(Ii-2) When b ≦ t <ab, the following formula 11-2 is derived.
[Formula 11-2] (tb) / (ab) ≦ k ≦ t / (ab)

(Ii-3) In the case of ab ≦ t ≦ a, since the right side of Equation 9-2 is 1 or more, the following Equation 11-3 is derived.
[Formula 11-3] (tb) / (ab) ≦ k ≦ 1
If k follows the probability density function p (k), then 0 ≦ k ≦ 1, and therefore p (k) = 0 when k <0 or k> 1.
Further, the following expression 12 is derived as a characteristic of the probability density function p (k).
[Formula 12]

  In order to determine the probability Pon energized at time t1 + t in each of the above cases (i) and (ii), the probability density function p (k) may be integrated within the previously obtained range of k. That is, the following formula 13 is derived when b ≧ 0.5a, and the following formula 14 is derived when b <0.5a.

[Formula 13]

[Formula 14]

By multiplying the energization probability Pon and the power consumption of the energized device, the expected value of the load power can be obtained. Therefore, the energization probability Pon represents the shape of the combined load when the number of energized devices (the number of loads) is a certain number or more and the combined power is considered to be equal to the expected value.
If Pon (t) is integrated over the entire time period, the expected value of the energization time is obtained, so that the value is b as shown in the following equation 15.

[Formula 15]

Further, when the energization adjustment coefficient k is given as a random number with a constant occurrence probability, the probability density function p (k) = 1, so that the probability of energization is equal to Pon (t). That is, if the energization adjustment coefficient k given by random numbers is used to determine the energization start times Ts of a large number of loads, the probability of energization is equal to Pon (t). The expected value of the magnitude of the combined power, that is, the combined power expected value We is expressed by Expression 16.
[Expression 16] We = load power × number of loads × Pon (t)

Here, the probability Pon of energization at time t1 + t when the energization adjustment coefficient k is given by a random number with a constant occurrence probability is obtained. That is, the following formula 18 is derived when b ≧ 0.5a, and the following formula 19 is derived when b <0.5a.
Note that the probability density function prnd (k) of a random number that takes a value of 0 or more and 1 or less becomes prnd (k) = 1 from the following equation 17 and prnd (k) = constant value.
[Formula 17]

[Formula 18]

[Formula 19]

FIG. 4 illustrates the probability Pon of energization when energization possible time a = 8 and energization time b is changed in the range of 1 to 7 hours. Referring to FIG. 4, when the ratio of the energization time b to the energization possible time a (b / a) is small, the probability of energization becomes a relatively flat shape, but around b / a = 0.5. It becomes a mountain shape and becomes flat again as b / a increases. As described in the example of FIG. 3, when b ≧ 0.5a, energization is always performed during a−b ≦ t ≦ b (the upper shaded portion in FIG. 3). This shows that it is difficult to adjust the shape of the current probability Pon regardless of how the probability density function p (k) is changed. In other words, when b / a is increased, the width a-b in which the energization start time Ts can be adjusted is shortened, and the adjustable range is reduced.
From the above, in order to adjust the shape of the combined power of all loads by adjusting the energization start time of each load, the probability density function p (k) of the energization adjustment coefficient k is matched to the load situation. It turns out that adjustment is necessary. Further, when the energization adjustment coefficient k is determined using a random number, a random number having a distribution according to the probability density function p (k) is obtained by multiplying the random number by a “random number modification function fn” to be described later, and a random number having a constant probability density. Need to be converted to

《Cumulative distribution function》
The cumulative distribution function P (k) of the probability density function p (k) will be described.
The cumulative distribution function P (k) of the probability density function p (k) can be expressed by the following equation 20 because p (k) = 0 when k <0 or k> 1. FIG. 5 illustrates this, and for example, the probability that the energization adjustment coefficient k is in the range of k to k + Δk is P (k + Δk) −P (k).
[Formula 20]

《Random number modifier function》
In this specification, the random number modification function refers to an inverse function of the cumulative distribution function P (k). In the example of FIG. 5, when a random number having a value of 0 or more and 1 or less is considered, the probability density function is prnd (k) = 1. The cumulative distribution function of random numbers is obtained by integration of a probability density function, and Prnd (k) = k, and the probability of occurrence of random numbers having a value between P (k1) and P (k1 + Δk) is Prnd (P (k1 + Δk)) − Prnd (P (k1)) = P (k1 + Δk) −P (k1). When a random number taking a value between P (k1) and P (k1 + Δk) is multiplied by the inverse function of the cumulative distribution function P (k), the random number is distributed between k1 and k1 + Δk. However, in order to uniquely determine k, P (k) needs to be a monotonically increasing function (p (k)> 0, where p (k) = 0 is allowed at one point). As a matter of course, the probability of occurrence of a number sequence distributed between k1 and k1 + Δk, which is obtained by multiplying a random number by the inverse function of P (k), is P (k1 + Δk) −P (k1). This relationship always holds as long as k1 and k1 + Δk are 0 or more and 1 or less. That is, the probability that a sequence obtained by multiplying a random number with a constant occurrence probability taking a value between 0 and 1 and the inverse function of the cumulative distribution function P (k) takes a value between k1 and k1 + Δk is P (k1 + Δk) −P (k1). This is because the cumulative distribution function of the obtained sequence is equal to P (k) by the definition of the cumulative distribution function. As described above, a random number having a constant occurrence probability can be converted into a sequence according to the cumulative distribution function P (k) or a probability density function p (k) obtained by its differentiation. An inverse function of the cumulative distribution function P (k) is a random number modification function.
As described above, the energization adjustment coefficient k is calculated by multiplying a random number modification function by a random number between 0 and 1, and the energization start time Ts is determined by the above equation 7, thereby determining the energization probability Pon (t). It can be controlled by the density function p (k) or the cumulative distribution function P (k).

The calculation procedure of the energization start time Ts using the random number modification function fn is expressed by the following formulas (i) to (iii).
(I) Generate a random number rnd of 0 or more and 1 or less. For the generated random number, a pseudo-random number is sufficient for practical use. Various methods for generating pseudo-random numbers have been proposed. Here, an example of a mixed congruential method with simple operations is shown.
a × p + q = a ′ In order to obtain a pseudorandom number of 4 digits, the last 2 to 5 digits are used.
For example, the first a = 5678, p = 678, q = 789.
5678 × 678 + 789 = 3850473 → 0.5047
5047 × 678 + 789 = 3422655 → 0.2265
2265 × 678 + 789 = 1536459 → 0.3645 (the same applies hereinafter)
Although the pseudorandom number rnd obtained by the mixed congruential method is 0 or more and less than 1, there is no practical problem.

(Ii) The energization adjustment coefficient k is determined by multiplying the pseudo-random number by the random number modification function fn. That is, the energization adjustment coefficient k can be expressed by the following equation 21.
[Formula 21]

(Iii) The energization start time Ts is determined by the above equation 8.

《Classification of random number modification function》
For the random number modification function, it is preferable to prepare a plurality of types in advance and select an optimal type according to the power demand situation. In the case of late-night power, for example, there may be a mountain-shaped demand curve in which demand increases or decreases around 4 o'clock, which is a negative peak, or around 1 o'clock at midnight and 6 o'clock in the early morning as shown in FIG. In some cases, a double-mountain demand curve with negative peaks will be drawn. In recent years, the introduction of the 2nd midnight power (1 to 6) for the purpose of leveling midnight demand has increased the demand for electricity at midnight due to the widespread use of electric water heaters energized during this time period. As leveling is achieved, the situation where the negative peak of power demand occurs twice is increasing. The shape of the power demand curve varies depending on the season. In order to realize such leveling corresponding to various power demands, it is effective to use a categorized random number modification function.
A procedure for obtaining the energization probability Pon (t) when various functions are introduced into the probability density function p (k) of the energization adjustment coefficient k will be described below.

(A) Maeyama type In order to shift energization to an earlier time zone, a distribution (hereinafter referred to as “maeyama type”) in which the probability distribution of k is biased toward a smaller k is considered. In order to realize the Maeyama type, a function of the following formula 22 is considered as a function to decrease by a power.
[Formula 22] p (k) = (r + 1) × (1-k) ^ r
r: positive integer

In Formula 22, the larger r is, the more it is biased forward. FIG. 7 shows a procedure for calculating the probability Pon (t) of energization according to the above equations 13 and 14 using the graph of the former mountain type probability density function p (k) and the function of equation 22.
Further, FIG. 8 shows the probability Pon (t) of energization when r = 5 and a = 8 hours. In this case as well, the effect of shifting to an earlier time zone increases as the ratio of b / a decreases, but in all cases the effect of shifting can be seen. In addition, although arbitrary positive integers can be selected for r, for example, when it is selected in the range of 3 to 10, it is likely that the general power demand can be handled (the same applies hereinafter).

(B) Hirayama type In order to shift the energization to a later time zone, a distribution that biases the probability distribution of k toward the larger k (hereinafter referred to as a “back mountain type”) is considered. In order to realize the rear mountain type, a function of the following Expression 23 is considered as a function to be increased by a power.
[Expression 23] p (k) = (r + 1) × k ^ r
r: positive integer In Equation 23, the larger r is, the later it is biased. FIG. 9 shows a procedure for calculating the probability Pon (t) energized by the above equations 13 and 14 using the graph of the back mountain type probability density function p (k) and the function of equation 23.
In addition, FIG. 10 shows the probability of energization Pon (t) when r = 5 and a = 8 hours. In this case as well, the effect of shifting to a later time zone increases as the ratio of b / a decreases, but the effect of shifting can be seen in all cases.

(C) Double Mountain Type In order to shift energization to an early time zone and a late time zone, consider a distribution that biases the probability distribution of k toward the smaller and larger k (hereinafter referred to as the “double mountain type”). The function of the following formula 24 is considered as a function obtained by averaging the front mountain type and the rear mountain type.
[Equation 24] p (k) = (r + 1) × ((1-k) ^ r + k ^ r) / 2
r: positive integer In formula 24, the larger r is, the more biased toward both ends. FIG. 11 shows a procedure for calculating the probability Pon (t) of energization by the above equations 13 and 14 using the graph of the double mountain type probability density function p (k) and the function of equation 24.
FIG. 12 shows the probability Pon (t) of energization when r = 5 and a = 8 hours. Also in this case, the shift effect to the both-end time zone increases as the ratio of b / a decreases, but the shift effect is seen in all cases. In the case of power demand as shown in FIG. 6, the Futama type is suitable.

(D) Uniform type When random numbers are used as they are, p (k) = 1. This is called a uniform type.

(E) Cumulative distribution function P (k) of uniform type, front mountain type, rear mountain type, and double mountain type
When the probability density function p (k) of each type is introduced into the equation 20, the following equation 25 is obtained. FIG. 13 shows the probability density function of each type when r = 5, and FIG. 14 shows the cumulative distribution function.
[Formula 25]
Uniform type P1 (k) = k
Maeyama type P2 (k) =-(1-k) ^ (r + 1)
Back mountain type P3 (k) = k ^ (r + 1)
Futatsuyama P4 (k) = (k ^ (r + 1)-(1-k) ^ (r + 1)) / 2

As described above, the random number modification function (random number is expressed as rnd) for controlling the energization start time Ts is an inverse function of the cumulative distribution function. Each type of random number modification function is represented by Equation 26 below. FIG. 15 shows each type of random number modification function in the case of r = 5.
[Formula 26]

Since some inverse functions other than the uniform type (k = rnd) cannot be expressed algebraically and the calculation capability of the control device for controlling the energization time is limited, the result of linear approximation (k = m × rnd + n) Shown below. Here, Table 1 shows a front mountain type random number modification function f2, Table 2 shows a rear mountain type random number modification function f3, and Table 3 shows a double mountain type random number modification function f4.

The typical random number modification function type has been described above, but the random number modification function fn may be any function as long as the following conditions are satisfied.
A) Probability density function p (k)
Since 0 ≦ k ≦ 1, p (k) = 0 when k <0 or k> 1.
[Formula 27]

B) Cumulative distribution function P (k)
[Formula 28]

C) Random number modification function fn (rnd) Inverse function of cumulative distribution function P (k) [Formula 29]

《Supplement for load leveling using random numbers》
In the method of adjusting the energization start time by using a random number and a random number modification function and performing load leveling, the load leveling effect obtained differs depending on the ratio b / a of the energizable time a and the energization time b. As / a increases, the load leveling effect decreases, and the shape of the resultant combined load also differs from the image expected from the shape of the probability density function p (k) of k. It is important to select a random number modification function suitable for the characteristics of each load by performing simulations as appropriate.
The random number modification function exemplified above is a constant function regardless of b / a. If the random number modification function is changed at the ratio of b / a, a greater leveling effect can be obtained, but the arithmetic processing becomes very complicated and cannot be handled by a CPU used in a normal control device. The main feature of this load leveling device is that it is possible to take measures with minimum equipment addition and low cost, and even with a simple random number modification function of the degree exemplified above, a considerable effect can be obtained. Is enough.
When b / a is small, the load leveling effect is large due to the large degree of freedom. However, the influence of the variation of the pseudo random number is increased accordingly, and the effect when the number is small tends to be considerably smaller than the expected value. Sufficient verification in simulation is necessary.
Since the power demand situation and load situation change, in order to respond flexibly to these changes, the control device is prepared with a plurality of random number modification functions as exemplified above and can be switched according to the situation. It is desirable to do so.

"Brief Summary"
FIG. 16 is a functional outline explanatory diagram of the load leveling means of the present invention described above, and the features are summarized as follows.
(1) Rather than using a large-scale control system that communicates with the central office, random numbers are used to control each device independently and probabilistically control the number of devices. It is an autonomous distributed control system that can be controlled in the form (expected value of probability theory).
(2) Since each device can independently determine the energization start time, the addition of operating devices does not affect the energization start time of other devices, and the control is very simple and the restrictions are extremely small.
(3) The device configuration is simply adding a switch and control device to the normal equipment power supply device. If the original device has a function to adjust the load energization time such as the use of midnight power, it can be handled only by adding control software. Functions can be realized with very simple additions. Therefore, equipment cost and operation cost are extremely low. (It is no longer necessary to shift the set value of the energizable time setting timer periodically.)
(4) Since the form of the combined load can be freely adjusted by introducing the random number modification function, it is possible to flexibly cope with various load forms and power supply and demand forms.

  Hereinafter, details of the present invention will be described by way of examples, but the present invention is not limited to the examples.

Example 1 relates to a method for leveling a charging load of an electric vehicle.
FIG. 17 shows an expected example of the charging time (normal charging) of the electric vehicle estimated from the daily travel distance of the passenger car. From FIG. 17, it can be seen that about 30% of the electric vehicles do not travel, so that charging is unnecessary, and there are many electric vehicles with a short charging time.

FIG. 18 shows a graph of the expected value of the charging load for normal charging of an electric vehicle when load leveling is performed using various random number modification functions. In FIG. 18, the charging load capacity is 3 kW / unit, the energization possible time is 8 hours, and the expected value of the charging load is obtained using Equation 16. Each graph in FIG. 18 means the following conditions.
No measures: Start charging all at the same time when energization is possible Energization control: Match charging completion to the end time when energization is uniform Uniform: Load leveling using random numbers using uniform random modification function Maeyama type: Maeyama type random number modification function Load leveling with random numbers using Horiyama type: Load leveling with random numbers using Gohyama type random number modification function Futama type: Load leveling with random numbers using Futama type random number modification function

It can be seen from FIG. 18 that a large peak occurs near the start time of energization in “no countermeasure”, and a large peak occurs in the vicinity of the end time of energization in “energization control”. In addition, when load leveling using various random numbers is performed, it can be seen that leveling in accordance with the characteristics of the random number modification function is performed because the charging time of electric vehicle ordinary charging is relatively short. More specifically, using a uniform random number modification function can reduce the load peak to about 1/4 compared to the case without countermeasures, and using the front and rear mountain random number modification functions can counter the load peak. It can be seen that it can be reduced to about ½ compared to the case without. In this case, it is not necessary to shift the set value of the energizable time setting timer that is periodically performed in a predetermined area unit. In addition, it can be seen that the Futama type random number modification function is suitable in the case of the late-night power demand situation as shown in FIG.
In addition, it can be said that the uniform type in which the maximum power is the smallest is desirable from the viewpoint of reducing the power receiving capacity of the user. If a customer with a large number of charged units such as a condominium takes such load leveling measures, the basic charge of electricity can be reduced.

The effect of the load leveling device of this embodiment is confirmed by simulation.
The load leveling device of this embodiment is applied to normal charging (3 kW / unit) of 50 electric self-propelled vehicles whose charging time follows the distribution of FIG. Each charging time, generated pseudo-random number, energization adjustment coefficient k obtained by multiplying the pseudo-random number with various random number modification functions, charging start time (the energizable start time is 0 hour, and the energizable time is 8 hours) FIG. 19 shows the composite load status for each random number modification function.
In FIG. 20, in the case of “no countermeasure”, the maximum power is 105 kW for a charging capacity of 150 kW (3 kW / unit × 50 units), but in the case of uniform load leveling, the maximum power is leveled to 42 kW. It has become.
Compared with FIG. 18 showing the expected value, the expected value of uniform leveling shown in FIG. 18 is 0.453 kW / unit, and the maximum power at 50 units is 23 kW. It can be seen that there are some deviations in the results. However, a correlation can be recognized between the shape of the graph of FIG. 20 and the shape of the graph of FIG. 18, and it is estimated that the deviation from the expected value is improved when the number of verified units is several hundred or more. The Therefore, the effect of the load leveling apparatus of the present embodiment could be confirmed by simulation.

Example 2 is related with the leveling method of the electricity supply load of an electric water heater.
FIG. 21 shows an example of energization time of an electric water heater using midnight power. From FIG. 21, it can be seen that the energization time of the electric water heater is considerably longer than the normal charging time of the electric vehicle, and the tendency is different.
FIG. 22 shows a graph of the expected value of the energization load of the electric water heater when load leveling is performed with various random number modification functions. Here, although the electric heater capacity | capacitance of an electric water heater is 4-6 kW normally, in FIG. 22, in order to compare with FIG. 18 which concerns on an electric vehicle, the electric heater capacity | capacitance is set to 3 kW / unit and the energization possible time is 8 hours. The expected value of the energization load was calculated using Formula 16 as in Example 1.
From FIG. 22, it can be seen that the electric water heater load has a long energization time, so that the load becomes a mountain shape in leveling using a uniform random number modification function. It can also be seen that the leveling effect is greatest when the Futama type random number modification function is used. Here, in order to further emphasize the Futama type, the power r of the random number modification function may be increased. Also in this embodiment, there is no need to shift the set value of the energizable time setting timer that is periodically performed in a predetermined area unit.

By the way, at present, there is a mixture of “no countermeasure” electric water heaters that start energization at the energizable start time and “energization control” type electric water heaters that match the hot water rising time with the energizable end time. For this reason, the effect is substantially the same as when load leveling is performed. However, many of the new electric water heaters are energized control types, and it is expected that the ratio of energized control types will increase in the future, and in that case, it will be necessary to newly take measures for load leveling. This point,
Since the energization control type electric water heater can realize the load leveling using the random number of this embodiment only by adding control software, the load leveling method of this embodiment is the energization control type electric water heater. It can be said that it is suitable for introduction into the system.

The effect of the load leveling device of this embodiment is confirmed by simulation.
This load leveling device is applied to 50 electric water heaters (3 kW / unit) whose energization time follows the distribution of FIG. Each charging time, generated pseudo-random number, energization adjustment coefficient k obtained by multiplying the pseudo-random number with various random number modification functions, charging start time (the energizable start time is 0 hour, and the energizable time is 8 hours) FIG. 23 shows the composite load status for each random number modification function.
In FIG. 24, in the case of “no countermeasure”, the maximum power is 147 kW with respect to the charging capacity of 150 kW (3 kW / unit × 50 units), but the maximum power is 84 kW when the double-mountain type load leveling is performed. Leveled.
Compared with FIG. 22 showing the expected value, the expected value of uniform leveling shown in FIG. 22 is 1.36 kW / unit, and 50 units is 68 kW. It can be seen that there is a gap. However, the shape of the graph of FIG. 24 and the shape of the graph of FIG. 22 can recognize a higher degree of correlation than in the first embodiment. Is estimated to be improved. Therefore, the effect of the load leveling apparatus of the present embodiment could be confirmed by simulation.

The present invention is applicable not only to electric water heaters and electric vehicles but also to power storage devices.
“Electric vehicle” refers to a vehicle in which a secondary battery is mounted on a vehicle such as an electric vehicle, an electric scooter, an electric wheelchair, etc., and the driving power is obtained using electric power charged in the secondary battery from the outside. That means.

Claims (12)

An energization load leveling method for energizing a plurality of energized devices in a specific power time period starting from time t1 and ending at time t1 + a,
For each energized device, the latest energization start time Tsm is calculated based on the energization time b of the energized device, and the energization start time Ts of the energized device is determined between the energization possible start time t1 and the energization start time Tsm. In this case, the energization adjustment coefficient k (0 ≦ k ≦ 1) is calculated by multiplying the random number rnd uniformly generated in a predetermined range by a preset random number modification function fn, and the energization start time Ts is calculated by the following formula A. An electrified load leveling method characterized by determining.
[Formula A] Ts = t1 + k × (ab) The random number modification function fn is an inverse function of the cumulative distribution function P (k) of the probability density function p (k), the probability density function is defined by the following formula B, and the cumulative distribution function P (k) is expressed by the following formula C. The energization load leveling method according to claim 1, wherein the energization load leveling method is defined.
[Formula B]
[Formula C]
The random number modification function fn selects one function from a previously created cumulative distribution function group including the following formula D, the following formula E, the following formula F, and / or the following formula G (where r is a positive integer). The energization load leveling method according to claim 2, wherein
[Formula D] P (k) = 1
[Formula E] P (k) = (r + 1) × (1-k) ^ r
[Formula F] P (k) = (r + 1) × k ^ r
[Formula G] P (k) = (r + 1) × ((1-k) ^ r + k ^ r) / 2   The energization load leveling according to claim 3, wherein when the specific power time zone is a midnight power time zone in which the second midnight power is set, the function of the formula G is selected as the random number modification function fn. Method.   The energized load leveling method according to claim 1, wherein the energized equipment includes an electric water heater and an electric vehicle.   The program which makes a computer implement the energization load leveling method in any one of Claims 1 thru | or 5. An autonomous energization load leveling device that is electrically connected to an energized device that is energized in a specific power time period that starts at time t1 and ends at time t1 + a, and that controls energization,
The energization load leveling device calculates the latest energization start time Tsm based on the energization time b of the energized device, and sets the energization start time Ts of the energized device between the energization possible start time t1 and the energization start time Tsm. In the determination, the energization adjustment coefficient k is calculated by multiplying a random number modification function fn set in advance with a random number rnd that is uniformly generated within a predetermined range, and the energization start time Ts is determined by the following equation A. An energizing load leveling device.
[Formula A] Ts = t1 + k × (ab) The random number modification function fn is an inverse function of the cumulative distribution function P (k) of the probability density function p (k), the probability density function is defined by the following formula B, and the cumulative distribution function P (k) is expressed by the following formula C. The energizing load leveling device according to claim 7, wherein the energizing load leveling device is defined.
[Formula B]
[Formula C]
A storage means for storing a cumulative distribution function group including a function of the following formula D, the following formula E, the following formula F and / or the following formula G (where r is a positive integer); 9. The energization load leveling device according to claim 8, wherein one function selected from among the functions is set as the random number modification function fn.
[Formula D] P (k) = 1
[Formula E] P (k) = (r + 1) × (1-k) ^ r
[Formula F] P (k) = (r + 1) × k ^ r
[Formula G] P (k) = (r + 1) × ((1-k) ^ r + k ^ r) / 2   10. The function of the formula G is set in the random number modification function fn when the specific power time zone is a midnight power time zone in which the second midnight power is set. The energizing load leveling device described.   The charging device for electric vehicles carrying the power supply load leveling apparatus in any one of Claims 7 thru | or 10.   An electric water heater on which the energization load leveling device according to any one of claims 7 to 10 is mounted.