In this paper, we consider the Jaulent-Miodek (JM) Equation [1]

We study the exact solutions of the JM Equation (1.1) by using Darboux transformation (DT), which is an effective method to get exact solutions from the trivial solutions of the nonlinear partial differential equations based on the Lax pairs [2] -[11] . As to the higher JM Equation, authors used several methods considering the travellling wave solutions [12] -[14] . For the solutions of the JM Equation (1.1), in [1] , the solitary wave solutions have been obtained by Darboux transformation. In this paper, we start from a different Lax pair to get some new exact solutions.

This paper is arranged as follows. Based on the Lax pair of the JM Equation (1.1), in Section 2, we deduce a basic DT of the JM Equation (1.1). In Section 3, from a trivial solution, we get solitary wave solutions of the JM Equation (1.1). Particularly, we obtain the bell-kink-type solitary wave solutions. We also get the elastic-inelastic- interaction coexistence phenomenon for the JM Equation (1.1). To the author’s best knowledge, this is a new phenomenon for the JM Equation (1.1).

We consisder the isospectral problem introduced in [15]

and the auxiliary spectral problem

From the zero curvature equation, we get the JM Equation (1.1).

We introduce a transformation

with

The Lax pair (2.1) and (2.2) is transformed into a new Lax pair

and

We suppose that

where, , , , , are functions of and.

Let and be two basic solutions of the Lax pair (2.1)

and (2.2). From (2.3), there exist constants such that

with

There are Equations and unknowns, , , , in (2.9). In order to determine these unknowns uniquely, we add another three Equations

The unknown in will be determined later.

From (2.8) and (2.9), we have

which means are roots of (note that is independent of).

Proposition 1. Let satisfy the Equation

Through the transformation (2.3) with (2.4), the isospectral problem (2.1) is transformed into (2.6) with

where are determined by (2.9) and (2.11).

Proof. Let and

It is easy to see that are (2N) th-order polynomials in, is a (2N-1)th-or- der polynomial in. By (2.1) and (2.10), we have Riccati Equation

Then all are roots of. Therefore we have

where

and are independent of. We can rewrite (2.17) as

By comparing the coefficients of, , with (2.11) and (2.13), we get

: (2.19)

: (2.20)

: (2.23)

From (2.21), (2.23) and (2.25), together with (2.11), (2.13), (2.14), (2.19), (2.20) and (2.24), we respectively get

Comparing with (2.4) and (2.18), we find that, and then and have the same form. □

Remark. When, supposing, DT is

Proposition 2. Let satisfy the Equation

where, , are defined by (2.9) and (2.11), and are defined by (2.14). Through the transformation (2.3) with (2.5), the auxiliary spectral problem (2.2) is transformed into (2.7) with (2.14).

To prove Proposition 2, we need to use Proposition 1 and the JM Equation (1.1), together with the help of the mathematical software (such as Mathematica). Although the idea of the proof for Proposition 2 is the same as Proposition 1, it is much more tedious and is omitted for brevity.

Since the transformation (2.3) with (2.14) transforms the Lax pair (2.1) and (2.2) into the same Lax pair (2.6) and (2.7), the transformation determined by (2.3) and (2.14) is called the DT of the Lax pair (2.1) and (2.2). Both the Lax pairs (2.1), (2.2) and (2.6), (2.7) obtain the JM Equation (1.1). Then, the transformation determined by (2.3) and (2.14) is also called the DT of the JM Equation (1.1).

In this section, by using of the above obtained DT, we get new solutions of the JM Equation (1.1).

For simplicity, taking, we get two basic solutions of the Lax pair (2.1) and (2.2)

with.

According to (2.10), we get

In the following, we discuss the two cases and.

1) For, from (2.9) and (2.11 ), we have

with. Then the exact solution of the JM Equation (1.1) is

with. This solution is similar with the solution in [11] .

As, this is a solitary wave solution where is a kink-type soliton and is a bell-kink-type soliton, i.e. this soliton is composed of a bell-type wave and a kink-type wave (see Figure 1).

2) For, from (2.9) and (2.11), we have

where

with

The exact solution of the JM Equation (1.1) is

When the parameters are suitably chosen, the solution (3.8) describes the elastic-inelastic-interaction coexistence phenomenon, i.e. the elastic and fission interactions coexist at the same time (see Figure 2).

In Figure 3, we can clearly find the interactions of the solitons. The solution is a solitary wave solution, where five kink-type solitons fuse into three kink-type solitons, i.e. K2 kink-type soliton and K4 kink-type

soliton are head-on interactions (this is an elastic interaction), K1 kink-type soliton, K3 kink-type soliton and K5 kink-type soliton fuse into K135 kink-type soliton (this is a inelastic interaction). The solution is a solitary wave solution, which is the same as, but the solitons are the bell-kink-type (see also Figure 3). This phenomenon has been described in the Whitham-Broer-Kaup shallow-water-wave model [16] . It seems to be new for the JM Equation.